Microfilmed 199S Information to Users

UMI MICROFILMED 199S INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleed through, substandard margins, and improper alignment can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with s m a ll overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

A Beil & Howell Information Company 300 North Zeeb Road. Ann Arbor. Ml 48106-1346 USA 313.-761-4700 800.521-0600

Order Number 0517044

Molecular and biochemical studies of RubisCO activation in Anabatna species

Li, Lih-Ann, Ph.D.

The Ohio State University, 1094

UMI 300 N. ZeebRd. Ann Arbor, MI 48106

MOLECULAR AND BIOCHEMICAL STUDIES OF RUBISCO

ACTIVATION IN ANABAENA SPECIES

DISSERTATION

Presented in Partial Fullfillment of the Requirements for

the Degree Doctor of Philosophy in the Graduate

School of The Ohio State University

Lih-Ann Li, B.S.

41 ♦ ♦ 41 4t

The Ohio State University

1994

Dissertation Committee: Approved by

C.J. Daniels

K.E. Kendrick

W.R. Strohl Adviser Department of Microbiology F.R, Tabita Copyright by Lih-Ann Li 1994 To My Parents

ii ACKNOWLEDGEMENTS

I thank Dr. F. Robert Tabita for his guidance and training during the course of my graduate study. I also thank my lab fellows, Janet Gibson, Te-Jin Chaw, James

Dubbs, Deane Falcone, Robert Ramagc and Katherine Tcrlesky for contributing their thoughts and technical assistance to this work. I would like to express appreciation to my dissertation committee members for their time and evaluation. I thank Dn. J. Ehlai,

P. Wolk, S. Nierzwicki-Bauer, S. Golden, B. Diner, K. Ohki and S.E. Steve ns, Jr. for sending us cyanobacteria! strains and Dr. D. Bryant for pointing out the potential close relationship of chloroplasts to heterocystous cyanobacteria, I am grateful to |Drs, W.L.

Ogren, G.W. Snyder and J.M. Wemeke for the spinach rca clone and the aiitiserum to spinach RubisCO activase and to Drs. J. Ehlai, P. Wolk and R. Haselkam for the

Anabaena conjugation system and rbcLr£cS*containing plasmids. I am also grateful to

Drs. R. Sayre, T.C. Huang and R.J. Mural for the psbA, nifH and p ic probes, respectively, to Dr. Y. Kawata for pKY206, to Dr. Darzins for pNOT19 anl pMOB3, and to Dr. A. Portis for CA1P. Finally, I wish to express sincere gratit ide to my parents, brother and sisters for their continuous encouragement and support. VITA

July 1, 1963 Bom in Taitung, Taiwan, Republic of China,

1985 B.S., Chinese Culture University, Taipei, Republic of China.

1986-1988 Graduate Research Associate, University of Texas, Austin, Texas, USA.

1989-1994 Graduate Research Associate, Ohio State University, Columbus, Ohio, USA.

PUBLICATIONS

Li L-A, Gibson JL and Tabita FR (1993) The Rubisco activase ( rca) gene is located downstream from rbcS in Anabaena sp. strain CA and is detected in other Anabaena/Nostoc strains. Plant Mol. Biol. 31: 753-764.

Tabita FR, Gibson JL, Falcone DL, Wang X, Li L-A, Read BA, Terlesky KC and Paoli GC (1993) Current studies on the molecular biology and biochemistry of C02 fixation in phototrophic bacteria. In: Murrell C and Kelly PD (eds.) Microbial Growth on Ct Compounds pp.469-479. Intercept Scientific Publication, Andover, UK.

FIELD OF STUDY

Major Field: Microbiology. TABLE OF CONTENTS

ACKNOWLEDGEMENTS...... iii

VITA...... iv

LIST OF TABLES...... vii

LIST OF FIGURES...... viii

CHAPTER PAGE

I. Overall Introduction

GENERAL BACKGROUND ON CYANOBACTERIA 1 CALVIN REDUCTIVE PENTOSE PHOSPHATE CYCLE 2 RUBISCO...... 5 PRK...... 11 ORGANIZATION OF CALVIN CYCLE GENES...... 12 SCOPE OF THIS DISSERTATION...... 14

II The RubisCO Activase (rca) Gene Is Located Downstream from rbcS in Anabaena sp. Strain CA and Is Detected in Other Anabaena/Nostoc Strains

INTRODUCTION...... 16 MATERIALS AND METHODS...... 18 RESULTS...... 23 DISCUSSION...... 35

v TABLE OF CONTENTS (continued)

CHAPTER PAGE

III Transcription Control of Ribulose Bisphosphate Carboxylase/Oxygenase Activasc and Adjacent Genes in Anabaena Species

INTRODUCTION...... 41 MATERIALS AND METHODS...... 43 RESULTS...... 47 DISCUSSION...... 66

IV Regulation of Anabaena RubisCO in vitro and in a RubisCO Activase Mutant

INTRODUCTION...... 73 MATERIALS AND METHODS...... 78 RESULTS...... 81 DISCUSSION...... 101

V Cyanobacterial Genes Encoding Phosphoribulokinase and Ribulose Bisphosphate Carboxylase/Oxygenase Are Not Closely Linked

INTRODUCTION...... 108 MATERIALS AND METHODS...... 110 RESULTS AND DISCUSSION...... 112

VI Concluding Remarks ...... 123

UST OF REFERENCES 129 LIST OF TABLES

TABLE PAGE

2.1 Bacterial strains and plasmids (used in CHAPTER II) ...... 19

4.1 Solubility and RubisCO activity of Anabaena sp. strain CA rbcL and rbcS gene products in Escherichia coli...... 85 LIST OF FIGURES

FIGURE PAGE

1.1 Calvin reductive pentose phosphate cycle and phosphorcspiratory oxidation of ribulose 1,5-bisphosphate (RuBP) ...... 3

1.2 Hypothetical mechanism of activase-mediated RubisCO activation 9

2.1 Southern hybridization localizes the rca and rbcS genes from Anabaena sp. strain CA within the same 4.3 kb EcoRI fragment 25

2.2 Restriction map of the fragment containing the Anabaena sp. strain CA rbcL, rbcS and rca genes ...... 26

2.3 Nucleotide and deduced amino acid sequence of the rca gene from Anabaena sp. strain CA ...... 28

2.4 Comparison of RubisCO activase sequences ...... 31

2.5 SDS gel electrophoresis (A) and Western immunoblot analysis (B) of Anabaena sp. strain CA RubisCO activase expressed in E. coli...... 34

2.6 Southern hybridization analysis of genomic DNA isolated from different cyanobacteria ...... 36

3.1 Genomic organization of C02-fixation genes in Anabaena strains 48

3.2 Nucleotide sequence and deduced amino acid sequences of the region between rbcL and rca in Anabaena sp. strain CA ...... 50

3.3 Potential transcriptional terminators for Anabaena rbcLXS and rca 54

3.4 Northern blot hybridization analysis of gene expression in Anabaena sp. strain CA ...... 57

3.5 Stability analysis of Anabaena sp. strain CA mRNA...... 60

viii 3.6 Influence of fructose metabolism on rbcL and rca transcription’ in Anabaena variabilis ...... 62

3.7 Primer extension analysis of the S' end of the Anabaena sp. strain CA rca mRNA and role of the 154 bp S' noncoding sequence in recombinant RubisCO activase synthesis ...... 63

4.1 Isomerization of RuBP during carboxylation ...... 75

4.2 Restriction maps of subclones used for expression of Anabaena sp. strain CA rbcL, rbcS and rca genes in Escherichia coli...... 83

4.3 (A) Time course of Anabaena sp. strain CA RubisCO activity (B) Time-dependent activation of RuBP-treated Anabaena sp. strain CA RubisCO...... 87

4.4 (A) SDS gel electrophoresis analysis of Anabaena sp. strain CA RubisCO activase purification (B) Time course of ATPasc activity by Anabaena sp. strain CA RubisCO activase...... 89

4.5 Inactivation of Anabaena variabilis rca gene ...... 93

4.6 Growth response of Anabaena variabilis wild type and rca mutant strain 353 ...... 97

4.7 Whole-cell RubisCO activity of Anabaena variabilis wild type and rca mutant strain 353 ...... 99

4.8 Time-dependent increase of RubisCO specific activity of Anabaena variabilis wild type and rca mutant strain 353 ...... 102

5.1 Southern blot hybridization using a spinach prk cDNA probe...... 113

5.2 Nucleotide and deduced amino acid sequences of the region of Anabaena strain CA genomic DNA hybridizing to a spinach prk probe ...... 115

5.3 (A) Nucleotide sequences with a high percentage of identity to spinach prk (B) N-terminal amino acid sequence comparison ...... 119

ix CHAPTER I

Overall Introduction

GENERAL BACKGROUND ON CYANOBACTERIA

Cyanobacteria (blue green algae) possess two photosystcms and perform oxygenic photosynthesis similar to land plants and eukaryotic algae. Although cyanobacteria are primarily regarded as typical photoautotrophic microorganisms, some strains can grow in the dark at the expense of organic substrates (204). In addition, cyanobacteria utilize various nitrogen compounds for growth and many species fix molecular nitrogen in the absence of combined nitrogen. Since N2 fixation is extremely sensitive to 0 2, N2-fixing cyanobacteria employ several distinct mechanisms to harmonize anoxygcnic N2 fixation with oxygenic photosynthesis.

One may generally categorize cyanobacteria within three different morphological groups: unicellular, nonheterocystous filamentous, and heterocystous filamentous strains.

Many unicellular and nonheterocystous filamentous cyanobacteria fix N2 in the dark, when oxygenic photosynthesis ceases. A few nonheterocystous strains can perform a type of facultative anoxygenic photosynthesis when 0 2-evolving photosystem II is suppressed by sulfide. Under anoxygenic photosynthetic conditions, nitrogenase synthesis is induced by a deficit of combined nitrogen; ATP and the reductants required for both C02 fixation and N2 fixation are derived frc m Photosystem I with sulfide serving as the electron donor. The segregation of oxygenic photosynthesis and N2 fixation is the predominant metabolic characteristic of heteroc) stous cyanobacteria, which fix N2 aerobically during daylight hours in the hete rocysts, specialized cells terminally differentiated from vegetative cells. Hclerocysts pc sscss a unique, thick envelope which some authors believe may serve as a barrier for 0 2 diffusion. These differentiated cells also lack oxygenic photosystcm II, contributing tc ' the anaerobic, low redox potential environment required for N2 fixation. While N2 fixation occurs only in hetcrocysts, photosynthetic C02 fixation occurs only in vegetat ve cells (45).

CALVIN REDUCTIVE PENTOSE PHOSPHATE CYCLE

Cyanobacteria assimilate and metabolize C02 primarily through the Calvin reductive pentose phosphate cycle, the major pathway for C 02 fixation in the biosphere.

Figure 1.1 illustrates the Calvin pathway by which C02 is reduced and subsequently metabolized. The Calvin cycle requires 9 molecules of ATP and 6 molecules of NADH to convert 3 molecules of C02 into glyceraldehyde-3-phosphate, indicating a tight link between the Calvin reductive pentose phosphate cycle and photophosphoiylation.

Phosphoribulokinase (PRK) and ribulose-1,5-bisphosphate(RuBP) carboxylase/oxygenase

(RubisCO) catalyze two successive reactions unique i to the Calvin cycle and they are thus considered to be the key enzymes of this pathway. The properties of each of the key enzymes will be discussed in the following section; . The rest of the reactions of the Fig. 1.1. Calvin reductive pentose phosphate cycle and photorcspiratory oxidation of ribulose 1,5-bisphosphatc (RuBP). Some enzymes functioning in the Calvin cycle and compounds involved in these two pathways are indicated. RuBPC/O (RubisCO in text), RuBP carboxylase/oxygenase; PRK, phosphoribulokinase; PGK, 3-phosphogIyceric acid (PGA) kinase; GPD, glyceraldchyde phosphate (PGAL) dehydrogenase; FBP, fructose 1,6-bisphosphatase; Ru5P, ribulose 5-phosphate; E-4-P, ciythrose 4-phosphate; Xyl-5-P, xylulose 5-phosphate; F-6-P, fructose 6-phosphate; F 1,6 di P, fructose 1,6-bisphosphate; S-7-P, sedoheptulose 7-phosphate; Se 1,7 di P, sedoheptulose 1,7-bisphosphate; DHAP, dihydroxyacetonc phosphate. [Taken from Tabita, 19S8 (179)].

3 4

COOK

ATP AOP

PDA

POK

COOP (jH/JP 7L~tT POAL UdIPOA CaO

fb p

HCOH

[HO

Fig. 1.1 Calvin cycle are employed to regenerate ribulose-5-phosphate (Ru5P), the substrate of

PRK, from 3-phosphoglycerate (PGA), the product of RubisCO-mcdiated carboxylation.

The enzymes catalyzing the other steps of the pathway are shared with the gluconeogenic and glycolytic pathways. In some organisms, the hydrolytic removal of the phosphate moiety from sedoheptulose 1,7-bisphosphate is catalyzed by an enzyme independent of

fructose 1,6-bisphosphatase. Therefore, sedoheptulose 1,7-bisphosphatase may be considered to be a third unique Calvin cycle enzyme in these organisms (179).

RUBISCO

Structure and assembly. RubisCO from most sources, including cyanobacteria, contains eight large subunits (L) and eight small subunits (S) with molecular weights of about 56,000 and 15,000, respectively (Type I RubisCO) (179). While the sites for activation, substrate binding, and catalysis all reside on the large subunit (115), the function of the small subunit is not clear. Recent studies, however, indicate that the small subunit is essential for stabilizing the holoenzyme and maximizing activity (4, 47,

96, 136, 169). The LgSg structure has been determined to be an Lg core, composed of four Li dimers around a fourfold axis, in association with two S4 tetramers at the top and bottom of the core. The hexadecameric enzyme contains eight active sites, each of which is formed from the C-terminal alpha/beta barrel domain of one subunit and the N- terminal domain of the second subunit of an L? dimer (3, 25, 87). Another type of

RubisCO (Type II) has been isolated from purple nonsulfur photosynthetic bacteria; the RJiodospiriltum rubrum (184) and Rhodobacter sphaeroidcs (59) enzymes are the best examples of this type. The Type II RubisCO is comprised of multiples of large subunits only; however, the primary structure of the Type II large subunit shows little similarity to the large subunit of the LgSg enzyme (58, 118).

In the chloroplasts, newly synthesized large subunits become associated with another protein before they arc incorporated into the holocnzymc (11). It was then suggested that this RubisCO binding protein is somehow involved in the assembly of the

LgSg RubisCO structure, yet this protein, of course, is not part of the holocnzymc (150).

Subsequently, it was shown that the RubisCO binding protein shares great homology in primary structure with theEscherichia coll chapcronin protein, GroEL, a protein which, along with GroES, is indispensable for the assembly of bacteriophage capsids (74).

When I^Sg RubisCO (Type I) enzymes or the L, (Type II) enzymes arc synthesized in

E. coli, the GroEL and GroES proteins arc required for the formation of the dimers from nascent unfolded large subunits and may participate in further assembly steps to form the recombinant LjSa protein as well (67,68). However, small subunits rapidly and spontaneously bind Lg cores to constitute the active LgSa enzymes in vitro without the assistance of other protein factors (6, 122).

Catalysis. RubisCO is the enzyme responsible for the actual C 02 assimilation step of the Calvin cycle. Through the mediation of RubisCO, C02 is added to the enedio! form of RuBP; the six carbon transition-state intermediate thus formed subsequently splits into two molecules of PGA (Fig. 1.1). In addition to carboxylation,

RubisCO also catalyzes the oxygenation of RuBP; this reaction yields one molecule each of PGA and 2-phosphoglycoIatc. The phosphoglycolate is then dcphosphotylated and excreted or further metabolized by cyanobacteria (Fig. 1.1) (180). RubisCO thus initiates two antagonistic metabolic pathways, photosynthetic carbon reduction and photorespiratory carbon oxidation. In addition, RubisCO catalyzes the formation of small quantities of pyruvate and the cncdiol of RuBP may be attacked by protons, yielding both isomerization and epimcrization products of RuBP (7, 41, 209). Since C 02 and 0 2 arc incompatible substrates for RubisCO, this bifunctional enzyme must have the ability to partition the carboxylase and oxygenase activities at any given ratio of C02 and 0 2. This capacity to discriminate between C02 and 0 2 is expressed in terms of the substrate specificity factor (VJCJVJKX where Vr and V, are the maximal velocities for carboxylation and oxygenation, respectively, and Kt and K„ are the Michaelis constants for carboxylation and oxygenation, respectively. Enzymes with a low specificity factor arc less efficient in carboxylation. It has been discovered that the specificity factor of cyanobacterial RubisCO enzymes is lower than that of the enzymes from higher plants and eukaryotic algae, and the lower carboxylation efficiency results mainly from the high

Kr of the cyanobacterial enzymes (83). However, an active inorganic carbon- concentrating process enables cyanobacteria to actively transport inorganic carbon and circumvent the high C02 requirement for RubisCO carboxylation (131).

Activation. The catalytic competence of RubisCO is dependent on the reversible binding of C02 to the e-amino group of a lysine residue at the active site of the large subunit (Lys-201 for plant large subunit) followed by the stabilization of the resultant carbamyl-enzyme intermediate by divalent cations (103). Mg2* is the preferred metal ion for carboxylase activity while the oxygenase activity favors Mn2+ (179). However,

C 02 concentration rather than metal concentration determines the initial rate of activation

(104). RubisCO activation occurs spontaneously in the presence of bicarbonate and Mg2* in vitro, but the K„ value for C02 in the spontaneous activation reaction is much higher than the C02 concentration in chloroplasts. A chloroplast protein, known as RubisCO activase, greatly increases the affinity of unactivatcd RubisCO for C02 in the light so that activation can occur at physiological C02 levels (128). Light and thylakoid membranes or ATP is required to support the activasc-catalyzed activation (121, 142, 197).

However, in vitro, RubisCO activase does not activate RubisCO without prior tight binding of sugar metabolite inhibitors, such as ribulose 1,5-bisphosphatc, xylulose 1,5- bisphosphate, 3-kctoarabinitoI 1,5-bisphosphatc, and 2-carboxyarabinitol 1-phosphate at the active sites of RubisCO. Therefore, an activation mechanism was proposed that considers RubisCO activase to interact directly with inhibitor-bound RubisCO, thus releasing the inhibitors from the active sites and increasing carbamylation; in order for

RubisCO activase to promote activation of RubisCO, ATP must be hydrolyzed (Fig. 1.2)

(127'»

Localization. The subcellular distribution of RubisCO has been determined for a number of cyanobacteria using immuno-electron microscopy. A variable, sometimes predominant, proportion of the RubisCO enzyme pool is associated with the polyhedral inclusion bodies, called carboxysomes, while other RubisCO molecules appear to be present in the cytoplasm. RubisCO is the most abundant protein species found in carboxysomes (27). Although both the cytoplasmic and carboxysomal RubisCO enzymes RubisCO Activase

E-RuBP ECM+RuBP

E-XuBP ECM+XuBP

ECM-KABP *»EC M +K A BP

ECM-CA1P »» ECM+CA1P

ATP ADP+P:

Fig. 1.2. Hypothetical mechanism of activasc-mcdiated RubisCO activation. Inhibitor-bound RubisCO is transformed into the active form (ECM) by an ATP-dcpendcnt reaction catalyzed by RubisCO activase, [Modified from Portis, 1992 (127)]. RuBp=ribulosc 1,5-bisphosphatc XuBPsxylulose 1,5-bisphosphosphate KABP=3-keloarabinitol 1,5-bisphosphatc CAlP=carboxyarabinitol 1-phosphate 10 arc capable of fixing C02 and appear to be virtually identical in all characteristics thus far examined, carboxysomes are important for active C02 fixation, especially under low

COj/high 0 2 conditions. This is based on several observations. First, carboxysomes arc found only in prokaryotes which grow autotrophically and fix C02 through the Calvin cycle under aerobic conditions. When carboxysome-posscssing autotrophs are grown with limiting levels of C02, the number of carboxysomes increases and a majority of the intracellular RubisCO is associated with carboxysomes. Cyanobacterial mutants with aberrant carboxysomes, or no carboxysomes, require high levels of C02 for growth despite the fact that these strains exhibit normal rates of accumulation of inorganic carbon

(9, 167). In addition, carboxysomes are absent in heterocysts, which are devoid of

RubisCO and do not catalyze RuBP-dependent photosynthetic C 02 fixation (27).

It has been postulated that carboxysomes are important components of the C02 concentration mechanism in cyanobacteria. Thus, C 02 is believed to be actively pumped into the cytoplasm in the form of HCO/ by a membrane transport system. Lack of cytoplasmic carbonic anhydrase and the impermeability of the cytoplasmic membrane to

HC03' drives HC03' to further diffuse into the carboxysomes, where low levels of carbonic anhydrase convert HC03* back to C02. The concentration of C 02 in this compartment is thus elevated to a level sufficient for RubisCO carboxylation (9, 139).

The absence of carbonic anhydrase in the cytoplasm has been indirectly demonstrated by expression of human carbonic anhydrase in the cytoplasm of Synechococcus sp. strain

PCC 7942. Induction of the recombinant carbonic anhydrase activity accelerates C02 efflux, diminishes the internal inorganic carbon pool, and results in a high level demand 11 for C 02 to support photosynthesis irrespective of the concentration of C02 supplied to the cells (130). Recently, carbonic anhydrase has been isolated from intact cyanobacteria! carboxysomes; the carboxysome-associated enzyme contains 78 to 96% of the total cellular carbonic anhydrase activity, as determined before the isolation of carboxysomes (132).

PRK

Structure. PRK has been isolated from a variety of organisms. The enzymes purified from plants and algae are composed of two identical subunits of 42,000 to

43.000 Mr. The enzymes of photosynthctic and chemosynthctic bacteria arc homooctamers with a subunit molecular weight ranging from 32,000 to 35,000 (179).

The structure of cyanobacterial PRK enzymes is distinct from the above two groups. In

Chlorogloeopsisfritschii, PRK was reported to be a hexamer of40,000 M, subunits (109) whereas theAnabaena cylindrica enzyme contains one 43,000 Mr polypeptide and one

26.000 Mr polypeptide, the latter of which was speculated to be a contaminant (161).

Similar to the quaternary structure, the deduced amino acid sequence of PRK from photosynthetic and chemosynthctic bacteria is very different from the plant and eukaryotic algal enzymes (181). On the other hand, the sequence of the cyanobacterial

Synechocystis sp. strain PCC 6803 prk gene shows significant homology to the plant and algal enzymes and is not homologous to the bacterial enzymes. Cys-16 and Cys-55 of spinach PRK, involved in ATP binding and activity regulation (114, 124), are well 12 conserved in the Synechocystis sequence at positions 19 and 41* respectively (175).

Despite the dissimilarity in the overall primary structure, homology is observed at substrate binding sites and with respect to an EGLH tetrapeptide of unknown function

(181).

Catalysis and regulation. PRK is involved in the regeneration of RuBP in the

Calvin cycle. This enzyme catalyzes the transfer of a phosphate group from ATP to

Ru5P (Fig. 1.1). PRK enzymes from many bacteria are regulated by NADH, whereas the kinases from plants, algae and cyanobacteria are activated after a cysteine disulfide bond is reduced via a light-dependent ferredoxin/thioredoxin system. Comparison of values reported in the literature indicates that the maximum specific activity (or kM) of the bacterial enzymes is considerably lower than that of PRK from oxygen-evolving photosynthetic organisms; the difference ranges from 3 to 77 fold (179).

ORGANIZATION OF CALVIN CYCLE GENES

In purple nonsulfur photosynthetic bacteria and chemosynthetic bacteria, the structural genes encoding Calvin cycle enzymes such as RubisCO, PRK, and fructose bisphosphatase are closely associated but the order of gene arrangement is different from species to species. Alcaligenes eutrophus has two duplicate clusters; one cluster is found on the chromosome and the other on the megaplasmid of this organism (77).

Rhodobacter sphaeroides also contains two sets of Calvin cycle genes (61) on different genetic elements (the large and small chromosomes) (176). In R. sphaeroides one cluster contains genes that encode the large and small subunits of the Type I LgSg RubisCO while the Type II L, RubisCO is localized within the second cluster on chromosome II

(61). RJiodobacter capsulatus also contains two forms of RubisCO but the genes for the

Type I RubisCO are not associated with any other Calvin cycle structural genes as in R. sphaeroides (56). All the structural genes of each bacterial cluster thus far examined are in the same orientation and expressed in a single transcriptional unit (56) with the exception of Rhodospirillum rubrum, where the gene encoding the Type II RubisCO, the only RubisCO found in R. rubrum, is oriented in a direction opposite to other Calvin cycle genes (44). In all instances, the product of the transcriptional regulator gene, cbbR, controls the level of transcription (56).

In higher plants and green algae, the genes for the large and small subunits of

RubisCO are chloroplast- and nuclear-encodcd, respectively. Small subunits are synthesized as a preprotein in the cytoplasm and subsequently processed and translocated into the chloroplasts, where they are assembled with large subunits to form the holoenzyme (43). By contrast, the large and small subunit genes in nonchlorophytic eukaryotic algae are both located on the chloroplast genome and linked in tandem as in autotrophic bacteria (15, 34, 36, 78, 79, 188), The linkage of the large and small subunit genes is also observed in cyanobacteria (2, 120, 165, 205, Chapter II).

However, the distance between the linked large and small subunit genes varies among species. Generally, the intergenic space is relatively small in autotrophic bacteria; for instance, only 11 bp are present between the large and small subunit genes in R. sphaeroides (58), while a putative open reading frame encoding a polypeptide of about 14 15,000 Mr exists in this region in some cyanobacterial strains (93, Chapter III). There has been no indication whether other Calvin cycle genes arc localized in the proximity of the RubisCO genes on the cyanobacterial chromosome or nonchlorophytic chloroplast genome.

SCOPE OF THIS DISSERTATION

Because of its agricultural and ecological importance, photosynlhetic C02 assimilation has stimulated extensive study for decades; this dissertation is one among these many investigations. The focus of this dissertation is directed to the regulation of the key Calvin cycle enzyme, RubisCO, in Anabaerta, specifically at the activation step in the catalytic process. In Chapter II, I report on the cloning of the structural genes encoding RubisCO and RubisCO activase from Anabaena sp. strain CA. The nucleotide and deduced amino acid sequences of the Anabaena CA RubisCO activase gene were determined and the physical location of the activase gene relative to the RubisCO large and small subunit genes was established. The distribution of the RubisCO activase gene in several different cyanobacteria was also examined in this chapter. Surprisingly, only the heterocystous cyanobacterial strains examined contain DNA hybridizing to an

Anabaena CA gene probe. Chapter III further explores the organization of RubisCO and

RubisCO activase genes in two other Anabaena strains, Anabaena variabilis and

Anabaena sp. strain PCC 7120. Chapter III also describes work on the regulation of transcription of the RubisCO and RubisCO activase genes under a variety of growth 15 conditions. In addition, possible cis-acting regulatory elements for these C02-fixation genes was addressed. In Chapter IV, studies on the expression and purification of recombinant RubisCO and RubisCO activase are described as well as the construction of a RubisCO activase mutant in Anabaena variabilis. Evidence from in vitro enzyme studies and physiological examination of the mutant suggests that RubisCO activation in

Anabaena is regulated differently from higher plants. This dissertation also briefly encompasses efforts to clone the gene encoding the other key enzyme, PRK.

Unfortunately, the sequence of the Anabaena CA DNA which hybridized to a plant PRK cDNA probe exhibited little homology to the plant PRK coding sequence. The nucleotide sequence, together with the potential open reading frame encoded by this DNA sequence, is presented in the last chapter, Chapter V. Chapter V also discusses the organization of the Calvin cycle genes in Anabaena. CHAPTER II

The RubisCO Activase {red) Gene Is Located Downstream from rbcS in Anabaena

sp. Strain CA and Is Detected in Other Anabaena/Nostoc Strains1

INTRODUCTION

Cyanobacteria assimilate C02 primarily through the Calvin reductive pentose phosphate cycle, using the enzyme ribulosc l,5*bisphosphate (RuBP) carboxylase/oxygenase (RubisCO) as the key catalyst (178). Regardless of the source from which RubisCO is obtained, in vitro enzymatic activity is absolutely dependent on spontaneous carbamyiation of Lys201 at the active site of the large subunit (103).

However, the noncarbamylated (or unactivated) form of RubisCO is often the predominant species found within the chloroplast or bacterial stroma (23, 91, 179). The unactivated enzyme exhibits a marked tendency to bind to RuBP (81) and other high affinity phosphorylated metabolites (82) with the net result that there is a severe retardation in the rate of formation of active enzyme from the enzyme-RuBP (E-R) or enzyme-metabolite (E-X) complex in vitro. In vivo, however, RubisCO may be substantially activated in the light, and part of this activation is thought to be due to the

lThis chapter has been published in Plant Mol. Biol.

16 enzyme RubisCO activase (128, 153, 170). RubisCO activase catalyzes the ATP- dependent activation (145, 174) of the E-R (144) or E-X complex (143) which has been shown to be energetically coupled to the light reactions of photosynthesis (21, 22).

Recently, the structural genes encoding the plant (151, 198) and Chlamydomonas (149)

RubisCO activase (red) was isolated and sequenced. Transcription of rca was found to be light-regulated (210), consistent with the proposed role of RubisCO activase in vivo.

Other than the fact that C02 fixation is light regulated, little is known of the regulation of C 02 fixation in cyanobacteria, procaryotes which catalyze a plant-type or oxygen-evolving photosynthesis (200). The rbcL and rbcS genes arc cotranscribcd in unicellular (166) and hctcrocystous filamentous cyanobacteria (120). The 5' and 3' sequences flanking the rbcLrbcS genes encode the proteins which are important for concentrating C 02 within the unicellular organism Synechococcus sp. strain PCC 7942

(49, 101). The prk gene, which encodes the other unique Calvin cycle enzyme, phosphoribulokinase, is not associated with rbcLrbcS (175, W.E. Borrias, personal communication) as it is in other bacteria (57, 85, 112). There is a paucity of information relative to the organization of genes important for C02 fixation in cyanobacteria, particularly in the heterocystous strains. In this study, we show that a gene encoding

RubisCO activase is located downstream from the rbcLrbcS genes of Anabaena sp. strain

CA (ATCC 33047); furthermore, two unknown open reading frames were found between rbcS and rca, highly suggestive of some functional organization. Homologous sequences were not readily detected in DNA from nonheterocystous filamentous or unicellular cyanobacterial strains when the Anabaena sp. strain CA rca gene was used as a probe 18 at high stringency.

MATERIALS AND METHODS

Organisms and plasmids. The organisms and plasmids used in this study arc listed in Table 2.1. Anabaena sp. strains CA (ATCC 33047) and IF were from this laboratory's culture collection (69, 171); Dr. K. Ohki sent us a culture of strain IF after our strain was lost. Anabaena variabilis sp. strain 29413F„, Anabaena sp. strain PCC

7120, and Anabaena sp. strain M-131 were provided by P. Wolk and J. Ehlai; Anabaena azollae by S. Nierzwicki-Bauer; Synechococcus sp. strain PCC 7942 from S. Golden;

Synecltocystis sp. strain PCC 6803 by B. Diner; and Agmenellum quadruplicatum strain

PR-6 (Synechococcus sp. strain PCC 7002), Coccochloris elabens sp. strain Di, Nostoc sp. strain Mac, Oscillatoria sp. strain JCM, and Oscillatoria sp. strain 3NT were from the collection of the late C. Van Baalen and were obtained from S.E. Stevens, Jr.

Media and growth conditions. Asp-2 medium (189) with or without 10 mM

NaNOj was used for the cultivation of all marine strains and Cg-10 medium (24) was used to culture fresh water strains. Anabaena sp. CA and IF, Synechococcus sp. PCC

7002, Nostoc sp. strain Mac, Coccochloris elabens sp. strain Di and Oscillatoria sp. strain 3NT were grown at 39°C, whereas the other strains were grown at 28°C. All cyanobacterial cultures were bubbled with 1 % COj in air and illuminated constantly. E. coli cells were routinely grown in Luria Broth (LB) medium at 37°C with the supplementation of the following antibiotics: kanamycin, 50 /zg/ml; tetracycline, 12.5 Table 2.1. Bacterial strains and plasmids

Strains and plasmids Relevant genotype or phenotype

Anabaena sp. strain CA heterocystous filamentous sp. strain IF heterocystous filamentous sp. strain PCC 7120 heterocystous filamentous sp. strain M131 heterocystous filamentous Anabaena azollae heterocystous filamentous Anabaena variablis sp. strain 29413FD heterocystous filamentous Cocochloris elabens sp. strain Di unicellular Nostoc sp. Mac heterocystous filamentous Oscillatoria sp. strain JCM nonheterocystous filamentous sp. strain 3 NT nonheterocystous filamentous Synechococcus sp. PCC 7002 unicellular sp. PCC 7942 unicellular Syneckocystis sp. PCC 6803 unicellular Escherichia coli HB101 hsds20 recA13 ara-14 proA2 lacYl galK2 rps20 Stri xyl-5 Escherichia coli JM109 recAl endAl gyrA96 thi hsdR17 supE44 relAl a (lac-proAB) {F1 traD36proAB lacPZaMIS} pVK102 Tcr Km' cos tra' Mob+ pK18 KmrtocZ(AM15) 20 /ig/ml; 5-bromo-4-chIoro-3-indoIyl-B-D-galactoside, 40 /ig/ml, and isopropyl-B-D- ihiogalactosidc (IPTG), 0.1 mM, were contained in LB agar plates for detection of inserts in plasmid pK18 (134).

DNA manipulations. Cyanobacterial DNA was isolated from cultures at late log phase using the sarkosyl-Iysate procedure (46), followed by CsCI gradient centrifugation.

Preparations of plasmid DNA from E, coli were made by the alkaline lysis method (13).

Restriction digestions, ligations, and mung bean nuclease treatments were carried out as recommended by GIBCO BRL life technologies, inc.. Transformations, agarose gel electrophoresis, nick translation labeling, and colony hybridizations were performed by standard techniques described by Maniatis et al. (106). Southern blot hybridizations were done by following the protocol supplied with Gene Screen Plus membranes. The stringency was lowered when the plant rca gene probe was used; hybridization temperatures were decreased to 50°C and washes were performed in 2 X SSC [0.3 M

NaCl, 0,03 M sodium citrate, pH 7.0] two times at room temperature for 5 min each, followed by two washes in 2 X SSC-1% sodium dodecyl sulfate (SDS) for 30 min at

50°C and two washes in 0.1 X SSC for 30 min at room temperature. When the

Anabaena sp. strain CA rca gene probe was used, conditions of high stringency included hybridization at 65 °C and two washes in 2 X SSC for 5 min at room temperature, followed by two washes in 2 X SSC-1 % SDS for 30 min at 65 °C, and two washes with

0.1 x SSC for 30 min at room temperature. Conditions of low stringency with the

Anabaena rca probe included hybridization at 55 °C, followed by two washes with 2 X

SSC at room temperature for 5 min, two washes at 2 X SSC-1% SDS for 30 min at 55°C, and two washes with 0.1 x SSC for 30 min at room temperature. The dideoxy chain termination method (156) was used for DNA sequencing.

Gene library construction. A gene library for Anabaena sp. CA was constructed essentially as described by Weaver and Tabita (196). The Anabaena sp. CA DNA was digested with ///ndlll over prescribed time intervals. The size range of fragments from different times of digestion was examined by agarose gel electrophoresis; fractions containing fragments of about 20 kb were pooled and the fragments ligated to //mdlll- digcstcd cosmid pVK102 (86). The recombinant DNA molecules were packaged in vitro with preparations of lambda protein particles (106) and then transfected into E. coli

HB101 (17).

Constructions of plasmids for DNA sequencing and expression. Figure 2.2 illustrates the strategy employed to obtain suitable subclones for DNA sequencing.

Derivatives of plasmid pK18 containing the 2.7 kb Hpal fragment inserted at the Smal site in either orientation, i.e., pK18::Hp2.7-l and pK18:: Hp2.7-2, were subjected to double digestion with ft/I and Xbal to generate a protected 3' protruding end and an

EroIII-accessible 3' recessive end, respectively. Nested ExoIII deletion clones were constructed following established protocols (8) with the exception that the DNA was treated with mung bean nuclease instead of SI nuclease after ExoIII digestion. To express \he Anabaena sp. strain CA rca gene in f. coli, clone pK18::Hp2.7-2, containing the rca gene in the correct orientation with respect to the lac promoter, was trimmed by removal of the portion of the insert between the EcoKV site and Sstl site (within the multiple cloning site), yielding plasmid pK18::EVl,7-2. All the subclones were 22 propagated in E. coli JM109 (207).

Protein expression and purification. An overnight culture of E. coli

JM109(pK18::EV1.7-2) was inoculated at a proportion of 1 ml inoculum to 100 ml fresh

LB medium containing kanamycin. After growth for 4 h, IPTG was added to a final concentration of 0.5 mM. After a 12 h induction period, the cells were washed twice and resuspended in ice-cold lysis buffer (50 mM Tris-HCl, pH 7.5, 10 mM EDTA, 5 mM fl-mercaptoethanol, 1 mM phcnylmcthylsulfonyl fluoride) and subsequently lyscd by sonication using fifteen 10-sec bursts at 4°C. The soluble protein fraction was obtained after first removing whole cells and debris by centrifugation at 12,000# for 10 min, followed by a final centrifugation at 120,000# for 1 h. The high-speed supernatant or soluble protein was fractionated with ammonium sulfate to 40% saturation at 4°C; recombinant RubisCO activase protein was found in the precipitate and was separated from other proteins after centrifugation at 20,000# for 20 min. Following overnight dialysis in 20 mM Tris-HCl buffer (pH 7.0), RubisCO activase was further purified by

Q-Sepharose FPLC chromatography, using a 0-0.5 M NaCl gradient to elute the protein

(163).

Western immunoblot analysis. Samples were fractionated in 11% SDS polyacrylamide gels (105) and transferred to nitrocellulose or polyvinylidene fluoride

(PVDF) membranes using a Bio-Rad (Richmond, CA) trans-blot SD cell and a modified carbonate transfer buffer containing 10 mM NaHCOj, 3 mM Na2C03 (pH 10,5) plus

0.0375% SDS. Immunoblots were performed according to the method of Blake et al.

(14). In this study, antibodies raised against spinach RubisCO activase (a gift from Dr. 23 W.E. Ogreti) were used as the primary antibody; the goat anti-mouse secondary antibodies were conjugated with alkaline phosphatase.

RESULTS

Cloning of Anabaena sp. strain CA rca gene. Anabaena sp. strain CA DNA fragments generated from partial //mdlll digestion, were ligated to cosmid vector pVK102, packaged in vitro, and transfected into E, coli HB101 in order to construct a genomic library. Cosmid clones containing the genes encoding the large and small subunits of Anabaena sp. strain CA RubisCO ( rbcL and rbcS, respectively) were isolated by means of colony hybridization, using a Synechococcus sp. strain PCC 6301

(Anacystis nidulans) rbcLrbcS heterologous probe. Positive Anabaena sp. strain CA clones were verified to contain rbcLrbcS inserts by sequencing, and both genes were organized similar to Synechococcus sp. strain 6301 (166) and Anabaena sp. strain PCC

7120 (120).

During the course of examining both genomic DNA and the positive rbcLrbcS clones for the presence of other Calvin cycle structural genes, a plant rca gene probe yielded strong hybridization in Southern blots to Anabaena sp. CA genomic DNA. Moreover, the rbcLrbcS cosmid clones were found to hybridize to the plant rca probe, suggesting that such a sequence might be localized near the rbcLrbcS genes. That this was indeed the case was shown by an experiment in which a 4.3 kb EcoRl fragment containing the rbcS gene from both the cosmid clone and genomic DNA was shown to hybridize to the 24 rca probe (Fig. 2.1).

Mapping and sequencing of the Anabaena sp. strain CA rca gene. To further localize the rca gene, the 4.3 kb EcoRl fragment was subcloned and hybridized to the plant rca gene after restriction with various enzymes. The rca gene was mapped to a 1.7 kb EcoRV-Hpal fragment within a 2.7 kb Hpal fragment (Fig. 2.2). A series of

EroIII-deleted clones of the 2.7 kb Hpal fragment were isolated in order to sequence both strands of the fragment. The results of DNA sequencing indicated that the

Anabaena sp. strain CA rca gene was located 1588 bp downstream of the rbcLrbcS gene cluster. There were two open reading frames, ORFI and ORFIt, encoding potential polypeptides with molecular weights of 7,000 and 32,000, respectively, within the 1588 bp fragment separating rbcS and rca. Each of the open reading frames was in the same transcriptional orientation (Fig. 2.2).

The nucleotide and deduced amino acid sequence of the Anabaena sp. strain CA rca gene2 (Fig. 2.3) indicated that the gene was composed of 1248 nucleotides, specifying 415 amino acids. The deduced amino acid sequence shared greater than 50% identity with the mature eukaryotic RubisCO activase enzymes (Fig. 2.4).

Expression of Anabaena sp. strain CA rca gene in E. coll. The rco-encoded

EcoRV-Hpal fragment was subcloned in expression vector pK18 (pK18::EV1.7-2), transformed into E. coli JM109, and a recombinant gene product synthesized after cells were induced with IPTG. A polypeptide with the approximate expected molecular weight

*The nucleotide sequence reported appears in the EMBL, GenBank and DDBJ Nucleotide Sequence Databases under the accession number X67942. 25

12 3 4

pVK102*>

Fig. 2.1. Southern hybridization localizes the rca and rbcS genes from Anabaena sp. strain CA within the same 4.3 kb £coRI fragment. TherbcLrbcS cosmid clone (lanes 1 and 3) and genomic DNA (lanes 2 and 4) of Anabaena sp. strain CA were digested with EcoRI and probed with spinach rca (lanes 1 and 2) and Synechococcus sp. strain 6301 rbcS (lanes 3 and 4). Fig. 2.2. Restriction map of the fragment containing the Anabaena sp. strain CA rbcL, rbcS and rca genes. The rca gene is located 1588 bp downstream of the rbcS gene and transcribed in the same direction as rbcLrbcS. The thick arrows above the map indicate the transcriptional orientation. The thin arrows below the map designate the extent and direction of sequence derived by using M13 primers. H, ///mffll; Hp, Hpal', Ev, EcoRV; E, Eco RI,

26 rbcL X S I II rca

H HpHpEV HpH Hp E H INI I 1 I

I 1

pK1 8::EV1.7 I------1

pK18::Hp2.7 I------1

Fig. 2.2

oto Fig. 2.3. Nucleotide and deduced amino acid sequence of the rca gene from Anabaena sp. strain CA. The start point of the coding region has been verified by N-terminal amino acid sequencing of the recombinant gene product synthesized in E. coli.

28 29

1 TATTAAGTTTTTTTTAGTTGAOCTACTGAGCAACCAGGAATTTTTACCAAAAGATTAGTA 61 AACATAAOATACCAAAAAAATTTAACATTAATTOAATACAACATTATCAGTGATAATTTT

1 2 1 ATGAGTTATTACATTGCTCCGCGATTTTTGGATAAACTAGCTGTTCACATCACTAAAAAC MBYYXAPRPLDKXtAVHXTKN

1 8 1 TTTTTAAATCTTCCTOaTGTaCGCGTTCCCTTOATTTTOGGTATTCACGOACGCAAAGGT FliHIiPGVRVPIiXLGXHQRlCG

2 4 1 GAAGGAAAAACTTTTCAATGTGAGTTAGCTTTCGAOAAAATGGGTGTGGAAGTTACACTC BGKTFQCBLAFBKMOVBVTL

3 0 1 ATCTCTGGCGGTGAATTGOAAAGTCCCGATGCAGGAGATCCAGCCAGGTTGATTCGOTTG ISOGBLBSPDAGDPARIiIRL

36 1 CGCTATCGGGAAACCGCAGAACTGATTAAAGTACGTOGTAAAATGTGCGTACTGATGATT RYRBTABZilKVRGKHCVLHI

4 2 1 AATGATTTAGATGCAGGTGCGGGTCGCTTTGATGAAGGCACTCAATATACTGTCAACACT NDLDAOAGRFDBGTQYTVHT

4 8 1 CAGTTGGTGAATGCCACACTGATGAATATTGCTGATAATCCCACAaATGTACAATTACCG QLVNATLKKIADHPTDVQLP

5 4 1 GGAAGTTACGATTCAACACCTCTAAGGCGTGTACCAATTATTGTCACA0GTAATGATTT7 OSYDBTPLRRVPIIVTGHDF

60 1 TCTACCCTCTACGCGCCGTTAATTCGAGATGGGCGGATGGAAAAATTTTATTGGGAACCT STZiYAPLIRDGRHBKFYHBP

6 6 1 CACCGCOATGAAAAGGTGGGAATTGTCGGTGGAATTTTTGCAGAAGATGGACTGTCACAA HRDBKVGXVGGXFABDGLSQ

7 2 1 CGTGACGTAGAAAAGTTAGTTGACAGTTTCCCCAATCAATCAATTGACTTTTTTAGTGCT RDVBKLVDSFPNQSIDFFSA

7 8 1 TTACGTTCCCGAATTTACGACGAACAAATCCGCGACTTCATTCATCAAGTGGGTTATGAG LRSRXYDBQIRDFIHQVGYB

8 4 1 AATGTCTCGTTGCGCGTGGTGAATAGTTTAGAAGGGCCACCAGCATTTAAAAAACCTGAT NVSLRVVNSX.BGFPAPKKPD

9 0 1 TTTACCTTGTCTCATTTGATTGAGTCTGCTAACTTCATGGTGGCTGAACAAAAACGCATC FTLSHLIBSANFMVABQKRI

96 1 GAAACTTCTCAATTGGTGGATGAATACAATCGTCTGAATCGAGGTAGAAGTTATCAACCT BTSQLVDBYNRLNRORSYQP

1 0 2 1 GCATCACCTGTTGCCGAAATAGCAACGAGTCAGCCGTCACCGAATGGAGTTAATCAACCC ABPVABIATSQPSPNOVNQP

1 0 8 1 CAGAGTGCGAGTCCCCATATCAGCCTAGAGACACAAGAACAAATCCGGCAAATCTTGGCT QBA8PHISLBTQBQIRQXLA

1 1 4 1 CAAGGTCATAAAATTACCTTTGAACACGTTGATAATCGCCGTTTTCGTACAGGTTCTTGG QGHKXTFBHVDNRRFRTG8W

Fig, 2.3 Fig. 2.3 (continued)

1201 CAAAOCTOCOOTACAATTCACOTCOXTOCTaAOTCTOATOCCATCTCaACTTTAOAATCT QBCOTlHVDABSDAISTLBfl 1261 TaTTTAOCAOAOTATAQOOOTOAOTATOTOCOCTTOOTAOOGATTOATCCOAAAOCCAAO CLAKYROSYVRLVQIDPKAK 1321 COACOOOTTOTOOAOACOATTATTCAaCOTCCaAATaOOACAAATTAOOAOaACTOACaT RRVVRTZZQRPMOTN* 1381 AACTOOTATCATACCCTAOCCCTTATCTTflAOaACTTTAOTCCTCAAOATAAQQOCTAOA 1441 OCCOCTTACTACAAACAAAAAATTAOTC Fig. 2.4. Comparison of RuBisCO activase sequences. The Anabaena sp. strain CA deduced amino acid sequence shares greater than 50% identity with the mature or presumptive mature eukaryotic RuBisCO activase sequences. The conserved amino acids are indicated by an asterisk. The two ATP-binding domains arc shaded. Ana, Anabaena; Ara, Arabidopsis; Chi, Chlamydomonas', Spi, Spinach. Gaps, marked by dots, are inserted for purposes of sequence alignment.

31 32 Ara AVKEDKQTDO DRWRGLAYDT SDDQQDITRG KGMVDSVFQA PHGT.GTHHA Spl AENEEKNT.. DKWAHLAKDF SDDQLDIRRG KGMVDSLFQA PADA.GTHVP Chi VAPSRKQMGR WRSIDAGVDA SDDQQDITRG REMVDDLFQG GFGAGGTBNA ** * * ** ** * •* ** Ana M.SYYIAPRF LDKLAVHITK NFLNLPGVRV Ara VLSSYEYVSQ GLRQYNLDNM MDGFYIAPAF MDKLWHITK NFLTLPNIKV Sp i IQSSFEYESQ GLRKYDIDNM LGDLYIAPAF MDKLWHITK NFLNLPNIKI Chi VLSSQEYLSQ S . . RASFNNI EDGFYISPAF LDKMTIHIAK NFMDLPKIKV

* * * * * * * * * * *** * * * * ***** ** *** Ana PLILGIHGKK GEGKTFQCEL AFEKMGVEVT LISGGELESP DAGDPARLIR Ara PLZLOSRI30K GQGKSFQCEL VMAKMGINP1 MMSAGELESG NAGEVRKLIR Spl PLXLGVWGGK GQGKSFQCEL VFAKLGINPI MMSAGELESG NAGEPAKLIR Chi PLILGIWGGK GQGRTFQCAL AYKKLGIAPI VMSAGELESG NAGEPAKLIR was/M **** * * ** * **** **** ****** * ***** * Ana LRYRETAELI KVRGKMCVLM INDLDAGAGR FDEGTQYTVN TQLVNATLMN Ara QRYREAADLI K.K<3KMCCLP INDLDAGAGR MGGTTQYTVN NQMVNATLMN Spl QRYREAADLI A.KGKMCALF XNDLEPGAGR MGGTTQYTVN NQMVNATLLN Chi TRYREASDII K.KGRMCSLF INDLDAGAGR MGDTTQYTVN NQMVNATLMN

****** *** ** * **** **** ***** *** ****** Ana IADNPTDVQL PGSYDSTPLR RVPIIVTGND FSTLYAPLIR DGRMEKFYHE Ara XADNPTNVQL PGMYNKEENA RVPIICTGND FSTLYGPLIL DGRMEKFLTG Spl IADNPTNVQL PGMYNXQDNA RVPIIVTGND FSTLYAPLIR DGRMEKFYWA Chi IADNPTNVQL PGVYKNEEIP RVPIVCTGND FSTLYAPLIR DGRMEKYYHN

* * * *** * *** ** ****** *** * ** Ana PHRDEKVGIV GGIFAEDGLS QRDVEKLVDS FPNQSIDFFS ALRSRIYDEQ Ara PTREDRIGV. WGIFRTDKIK DEDIVTLVDQ FPGQSIDFFG ALRARVYDDE Spl PTREDRIGVC TGIFRTDKVP AEHWKLVDA FPGQSIDFFG ALRARVYDDE Chi PTREDRIGVC HGIFQEDNVQ RREVENLVDT FPGQSIDFFG ALRARVYDDM * * * * * * * * * Ana IRDFIHQVGY ENVSLRWNS LEGPPAFKKP DFTLSHLIES ANFMVAEQKR Ara VRKFVESLGV EKIGKRLVNS REGPPVFEQP EMTYEKLMEY GNMLVMEQEN Spi VRKHVNSVGV DNVGRKLVNS KDGPPVFEQP EMTLQKLMEY GNMLVQEQEN Chi VRQWITDTGV DKIGQQLVNA RQKVAM.PKV SMDLNVLIKY GKSLVDEQEN * * * Ana IETSQLVDEY NRLNRGRSYQ PASPVAEIAT SQPSPNGVNQ P...QSASPH Ara VKRVQLAETY ...... LSQAAL GDANADAIGR GTFYGKGAHE Spi VKRVQLADQY ...... MSSAAL GDANKDAIDR GTFFGKAAQQ Chi VKRVQLADAY ...... LSGAEL AGHGGSSLPE A..YSR Ana ISLETQEQIR QILAQGHKIT FEBVDNRRFR TGSWQSCGTI HVDAESDAIS Ara VNLPVPEGCT DPVAENFDPT ARSDDGTCVY NF Spi VSLPVAQGCT DPEAKNYDPT ARSDDGSCTY NL Ana TLESCLAEYR GEYVRLVGID PKAKRRWET IIQRPNGTN

Fig. 2.4 33 (Mr of about 49,000) was detected on SDS polyacrylamide gels only in induced cell extracts of strain JM109(pK18::EV1.7-2), but not in strain JM109(pK18) (Fig. 2.SA),

The pK18::EV1.7-2 dependent polypeptide crossreacted with antibodies raised against spinach RubisCO activase (198) in Western immunoblots (Fig. 2.5B). A second minor cross-reacting protein of lower molecular weight was also observed.

To further characterize the recombinant rca product, the protein was fractionated by ammonium sulfate precipitation and Q-Sepharose FPLC chromatography (163). The polypeptide eluted as a major component between 0.3 to 0.5 M NaCl from the

Q-Sepharose column. A fraction containing the expression product was subsequently clectrophorcscd in a SDS polyacrylamide gel in order to separate the recombinant protein from minor contaminants. This preparation was blotted to a PVDF membrane and the putative rca gene product extracted and prepared for N-terminal amino acid sequencing

(79). The N-terminal sequence was determined to be -SYYIAPRFLDKLAV-

1TKNFLNLPGV-V, thus confirming the start site deduced from the nucleotide sequence

(Fig. 2.3) and establishing that the polypeptide purified from the cell extract of E. coli

JM109(pK18::EV1.7-2) was the product of the Anabaena sp. strain CA RubisCO activase gene.

Hybridization of Anabaena sp. strain CA RubisCO activase gene to DNA from other cyanobacteria. Genomic DNA isolated from a variety of cyanobacteria was hybridized to the nick translated HP-labeled Anabaena sp. strain CA rca gene under conditions of high stringency. Thus far, genomic DNA from all heterocystous

Anabaena/Nostoc strains examined exhibited strong hybridization to the Anabaena sp. 34

Fig. 2.5. SDS gel electrophoresis (A) and Western immunoblot analysis (B) of Anabaena sp. strain CA RubisCO activase expressed in E. cob. The membrane-free extracts were prepared from IPTG-induced E. coli cells harboring vector pK18 (lane 1) and recombinant plasmid pK18::Ev!.7-2 (lane 2). The blot was probed with antibodies raised against spinach RuBisCO activase. Molecular weight standards, from top to bottom, are phosphorylase B(Afr 97,400), bovine serum albumin (Mr 66,200); ovalbumin (A/r 45,000); and carbonic anhydrase (Mr 31,000) and are found to the left of lane 1 in (A). 35 strain CA rca probe, while genomic DNA from unicellular and nonheterocystous filamentous cyanobacteria did not hybridize at high stringency (Fig. 2.6). When the spinach and Anabaena sp. strain CA rca genes were used as heterologous probes at low stringency, both probes hybridized to a discrete and clear 4.3 kb £coRl fragment of

Anabaena CA DNA, but reacted poorly with restricted Synechococcus 7942 and 7002

DNA, as a diffuse and barely resolved smear, that was difficult to attribute to specific hybridization (data not shown).

DISCUSSION

The activation of RubisCO by RubisCO activase is ATP-dependent. The two

ATP-binding domains identified in plant RubisCO activase enzymes (198) are highly conserved in the Anabaena sp. strain CA enzyme (amino acid residues 37 to 44 and 93 to 102) and the lysine residue essential for ATP binding in the G-G-GKS/T glycine rich consensus domain (163) is also conserved (residue 43). By contrast, the amino acid sequence homology sharply differs at the carboxy-terminus. The Anabaena RubisCO activase possesses additional residues at the carboxy-terminus but lacks much of the amino-terminal region present in plant and algal RubisCO activase. Thus, the additional carboxy-terminal sequence compensates for the truncated amino-terminus, yielding a molecular weight similar to the mature plant and algal enzymes. Besides specificity for different RubisCO enzymes, differences in the primary structure of RubisCO activase might reflect divergent regulation of C02 fixation, since unactivated cyanobacterial Fig. 2.6. Southern hybridization analysis of genomic DNA isolated from different cyanobacteria. An Attabaena sp. CA rca gene was used as a heterologous probe under conditions of high stringency. The probe was hybridized to DNA isolated from the following organisms: (A) lane 1, Synechococcus sp. strain PCC 7002; lane 2, Anabaena sp. strain IF; lane 3, Anabaena variabilis strain 29413FD\ lane 4, Anabaena sp. PCC 7120; lane 5, Anabaena sp. strain M-131. All preparations were digested with Hindlll. (B) lanes 1, 6, 11, Anabaena sp. strain CA; lane 2, Anabaena azoltae; lane 3, Synechococcus sp. strain PCC 7942; lane 4, Coccochloris elabens strain Di; lane 7, Nostoc sp. strain Mac; lane 8, Oscillatoria sp. strain JCM; lane 9, Oscillatoria sp. strain 3NT; lane 10, Synechocystis sp. strain PCC 6803. All preparations were digested with EcoRl.

36 !

S 6 7 8 9 I f 11

Fig. 2.6 38 RubisCO (E) is apparently not associated with RuBP (R) to form an E-R complex (95).

Current studies are thus directed at elucidating the physiological role of cyanobactcrial

RubisCO activase.

There is no detectable Shinc-Dalgamo sequence (165) in the immediate upstream region of the Anabaena sp. strain CA rca gene although cyanobacteria have been thought to have E. coli-l ike ribosome binding sites (16). This docs not appear to be unique in

Anabaena (31, 111, 120). In addition, nucleotide sequence analyses revealed that rbcLrbcS, rca, as well as the two intcrgcnic unknown open reading frames of Anabaena sp. strain CA were all in the same transcriptional orientation. It is not clear whether all these genes arc members of the same operon. However, preliminary Northern blot hybridization results showed that Anabaena sp. strain CA cells grown under constant illumination did not accumulate detectable amounts of rca message, while transcripts hybridizing to a strain CA rbcL probe were just large enough to encode the rbcL and rbcS genes (data not shown).

Because cyanobacteria, algae, and plants exhibit similar mechanisms to generate energy through photosynthesis and Anabaena sp. strain CA was found to contain homologous RubisCO activase, it was expected that the rca gene might be present in all cyanobacteria. Surprisingly, Southern blot hybridization analysis did not support this hypothesis. Indeed, the hybridization results indicated that genomic DNA from a variety of unicellular and nonheterocystous filamentous cyanobacteria did not readily recognize the Anabaena sp. strain CA rca probe under the conditions of high and low stringency reported here. Certainly these results do not exclude the possibility of some form of 39 RubisCO activase that is only poorly homologous to plant and hetcrocytous cyanobacteria! rca. Recent studies with Synechococcus sp. strain 7942 have not uncovered any rca-Wkc genetic information downstream from rbcS in this organism

(101).

Cyanobacteria have long been suggested to be the endosymbiotic progenitors of chloroplasts (108). Since C02 fixation is an essential function for cyanobacteria and chloroplasts, RubisCO activase may serve as an alternative to understand the phylogenetic relationship between cyanobacteria and chloroplasts; although the rca gene is nuclear-cncodcd in plants, there is evidence for gene transfer from chloroplast to nucleus (10, 51), According to the hybridization analyses, heterocystous cyanobacteria seem to be phylogenctically closer to chloroplasts than arc unicellular and nonhcterocystous filamentous cyanobacteria. Although we have examined only a limited number of cyanobacteria! strains at this time, our results are consistent with the hypothesis of Bryant and Stirewalt (18) suggesting that chloroplasts might be derived from an ancient Nostoc species, based on the broad symbiotic host range of Nostoc (200),

16S rRNA sequences (63) and the organization as well as the sequence of the ATP synthase gene cluster (29, 30, H I). Despite conflicting views about the origin of chloroplasts, the prochlorophytes (cyanobacteria producing chlorophyll b but no phycobilins) (97, 187) and the cyanelles (photosynthetic organelles containing phycobilisomes, carboxysomes and peptidoglycan envelope) (1, 195) are two other main lineages of chloroplast evolution. Heterocystous cyanobacteria were separately found to be more closely related to prochlorophytes and to cyanelles than unicellular 40 cyanobacteria by analysis of 16S rRNA (63, 160). However, the acquisition of

RubisCO activase could simply be the result of convergent evolution amongst heterocystous cyanobacteria and plants due to some sort of pressure. The results we have collected at this time, with a limited selection of cyanobactcrial strains, certainly suggests that additional studies with a more complete array of species is warranted.

There are many questions that remained unanswered. Paramount among these is elucidation of the actual role RubisCO activase plays in Anabaena, particularly since unactivatcd cyanobactcrial RubisCO docs not bind RuBP tightly. It may be that RubisCO activase in heterocystous cyanobacteria catalyzes an enzyme-mediated removal of some phosphate inhibitor (143, 144). In addition, it is not clear under what circumstances the

Anabaena rca gene is transcribed, since there appear to be no detectable transcripts under constant illumination. Finally, it is necessary to determine the role of the two unknown open reading frames upstream from rca and downstream from rbcS and whether they are involved in the regulation of C02 fixation. CHAPTER in

Transcription Control of Ribulosc Blsphosphatc Carboxylasc/Oxygcnasc Activase

and Adjacent Genes in Anabaena Species1

INTRODUCTION

Ribulosc l,5*bisphosphate (RuBP) carboxylasc/oxygcnasc (RubisCO) regulates photosynthetic and photorespiratory carbon metabolism by catalyzing either the carboxylation or oxygenation of RuBP (72). Before catalysis occurs, the enzyme must first be activated via the binding of C02 to the e-amino group of Lys-201 at the active site on the large subunit; this carbamylated enzyme intermediate is stabilized by divalent cations (103). In vivo, however, it is apparent that other factors, particularly light, are important for the regulation of RubisCO activity (80). In this regard, the isolation and characterization of a mutant of Arabidopsis thaliana which contained a defect in the light activation of RubisCO in vivo (170) led to the demonstration that RubisCO activation is catalyzed by another protein, RubisCO activase (153). RubisCO activity may also be retarded through its association with various phosphorylated sugar metabolites that accumulate in the celt. For instance, the substrate, RuBP, binds the unactivated form of

^his chapter has been accepted for publication in J. Bacteriol. 42 RubisCO from most sources and prevents the subsequent activation by C02 and Mg2*

(26, 81, 90). Also, the compound 2-carboxyarabinitol l*phosphate (CA1P), an analog of the transition state intermediate formed during RuBP carboxylation, competes with

RuBP for the catalytic site of activated RubisCO. CA1P also mediates part of the light regulation of C02 fixation in some plants since this compound is produced noctumally and degraded in the light (162). Recently, it was discovered that the characteristic

"failover" in the RubisCO activity rate curve is caused by the binding of compounds formed as a result of substrate isomerization and epimerization, such as xylulose 1,5- bisphosphatc (XuBP) and 3-ketoarabinitol 1,5-bisphosphate (KABP). Each of these compounds is formed by mistaken protonation of the cnediolatc of RuBP, and depending on pH, RubisCO activity will be progressively inhibited by the binding of XuBP to the unactivatcd form of the enzyme, or of KABP to the activated form of the enzyme (41,

206, 209). Although the mechanism for RubisCO activase-mcdiated activation is not clear, RubisCO activase is thought to facilitate the formation of activated RubisCO, while consuming ATP, partly through its ability to catalyze the removal of enzyme-bound metabolic inhibitors such as RuBP, CA1P, XuBP, and KABP (126, 143, 144, 174).

To date, purified RubisCO activase has been isolated only from higher plants and green algae (155), although there is some indication that a similar protein may be present in the cytosol of cyanobacteria (48). While the activase gene (rca) was recently detected and isolated from heterocystous cyanobacteria (Anabaena/Nostoc species), in particular the marine strain Anabaena sp. strain CA, there was no clear indication that homologous sequences were found in unicellular and nonheterocystous filamentous cyanobacteria (98), 43 perhaps indicating that the activase-] ike protein previously observed in these organisms

(48) might be quite different from the product of the rca gene of heterocystous species.

Interestingly, the deduced amino acid sequence of the Anabaena sp. strain CA rca gene showed many differences and some similarities to the plant and algal RubisCO activase

(98). However, neither the Anabaena nor the Synechococcus RubisCO exhibits failover and these enzymes are not inhibited by RuBP (S, 95, 100). Because there is no obvious need for activase to modulate the above aspects of RubisCO catalysis, the recent isolation of the gene that encodes activase in heterocystous cyanobacteria (98) prompted questions as to the actual role of activase in these organisms. Obviously this will not be resolved until there is more knowledge of both the genetics and biochemistry of activase in

Anabaena species. As a first step towards gaining an understanding of the importance and regulation of RubisCO activase in cyanobacteria, the current investigation on rca gene expression control was undertaken.

MATERIALS AND METHODS

Cyanobacterial growth conditions. Each of the organisms used in this investigation was cultured under conditions previously described (98). For Northern

RNA analysis, Anabaena variabtlis was grown under photoautotrophic conditions in BG-

11 medium (141) containing 1.5 g of NaNOj per liter and 10 mM N-(2-hydroxyethyl)- piperazine-N'-(3-propanesulfonic acid) (pH 8.0) and continuously bubbled with air; 10 mM fructose was added for heterotrophic growth. 44 Nucleic acid preparations and DNA sequencing. Cyanobactcrial DNA and

Escherichia coli plasmids were isolated as previously described (98). Anabaena RNA was purified from 100 ml of mid logarithmic phase cultures, to which 5 ml of ice cold stop solution (200 mM Tris-HCI, 20 mM EDTA, 20 mM sodium azide, [pH8.0]) was immediately added to terminate cell metabolism. The RNA was further purified by the procedure of Golden et at, (64).

Plasmid pAn600, containing the rbcLrbcS genes from Anabaena sp. strain PCC

7120 (31) and plasmid 21C1, containing the A. variabilis rbcLrbcS genes (75), were obtained from R. Haselkom and C.P. Wolk, respectively. Vectors pUC18/19 and pK18/19 were used to clone restriction fragments of Anabaena DNA and their exonuclease IH-digested derivatives for DNA sequencing. The dideoxy chain termination sequencing method (156) was used to sequence double stranded DNA templates using the

U.S. Biochemical Sequenase version 2.0 kit and M13 primers.

Nucleic add hybridization. Southern and Northern blot hybridizations were performed according to the supplemented protocols supplied with GeneScreen Plus membranes (Du Pont NEN Research Products). When the presence of open reading frames 1 and 2 (ORF1 and ORF2) was examined in A, variabilis and Anabaena sp. strain

PCC 7120, along with genomic DNA from several other cyanobacteria, the temperature used for hybridization and washing was lowered to 55°C and the concentration of sodium dodecyl sulfate (SDS) in the second-step washes was decreased to 0.1%. For Northern blots, RNA (15 pg of each sample) was denatured and fractionated on 1.5% formaldehyde-agarose gels (106). After transfer of RNA to a GeneScreen Plus 45 membrane, the blot was baked at 80°C for 2 h in order to reverse the formaldehyde reaction. Hybridizations were carried out in aqueous solutions containing 10% dextran sulfate, 1% SDS, and more than 500 /xg of denatured salmon sperm DNA per ml. A solution of 2X SSC (IX SSC is 150 mM sodium chloride plus 15 mM sodium citrate)-

2% SDS was used in the second-step washes. The appropriate sizes of the hybridizing fragments were determined by their migration in gels relative to a molecular weight ladder. Probes for hybridization were derived from the 0.9-kb WndlH fragment ( rbcL ),

0.5-kb EcoRI-Z/fVidlH fragment (rbcS), 1.0-kb Hpal-EcoRV fragment (ORF1-ORF2), and

0.7-kb Dral fragment (rca) of Anabaena sp. strain CA (ATCC 33047) DNA (Fig. 3.1 A).

The 1.8-kb Wmdlll inserts of pBM5 (117) and pAnl54.3 (140) were separately used as psbA and niJH probes, respectively. All the probes were labeled by means of nick translation (106).

Transcript stability analysis. Two Anabaena sp. strain CA cultures were grown in medium free of combined nitrogen and bubbled with 1 % C02 in air under constant illumination to mid-log phase. RNA synthesis in each culture was then inhibited by the addition of rifampcin (60 /xg/ml). One culture was kept in the light, whereas the second culture was shifted to the dark immediately after the addition of rifampcin. During the course of the experiment, rifampcin was replenished in 60 /xg/ml per hour. At each indicated time point, 100 ml of cells was withdrawn and RNA was isolated; 15 /xg of

RNA from each time point was subjected to Northern blot hybridization analysis.

Radioactivity was quantitated using a Betascope 603 blot analyzer (Betagen). Areas containing discrete hybridization bands and significant positive signals derived from 46 degraded transcription products in the sample before rifampcin addition were selected, the radioactivity from the selected areas was summed, and background radioactivity from areas of the same size subsequently was substractcd. The net radioactivity from equivalent areas in other lanes of the blot was calculated in the same manner. Half lives of the various transcripts were determined from a semi-log plot relating the loss of radioactivity over time.

Primer extension analysis. A 20-mer oligonucleotide (5'-

CAAGGGAACGCGCACACCAG-3'). complementary to the +94 to +74 region of the

Anabaena sp. strain CA rca coding sequence, was end labeled with [y-i2P]ATP. About

106 cpm of labeled primers was mixed with 200 fig of Anabaena sp. strain CA RNA, previously isolated from a culture grown in nitrogen-free ASP-2 medium (189) bubbled with 1% C02 in air. Primer extension analysis was performed according to published protocols (8).

Recombinant RubisCO activase synthesis. Plasmids pK18, pK18::EV1.7-2, containing the Anabaena sp. strain CA rca gene in the correct orientation with respect to the lac promoter of pK18 (98), and pK18::EV1.7-l, containing the rca gene oriented opposite to the lac promoter of pK18, were transformed into E. coli JM109, and recombinant gene products were synthesized as previously described (98).

Nucleotide sequence accession number. The nucleotide sequence reported in this paper has been submitted to the GenBank/EMBL data bank under accession number

U05590. 47

RESULTS

Nucleotide sequence and organization of the structural genes. An earlier study

(98) established that the rca gene of Anabaena sp. strain CA was situated downstream from the structural genes encoding the large ( rbcL ) and small (rbcS) subunits of RubisCO

(Fig. 3.1). To gain a better understanding of how the expression of these genes is controlled, the nucleotide and deduced amino acid sequences for the region between the

Anabaena sp. strain CA rbcL and rca genes were determined (Fig. 3.2). These analyses yielded both predicted and unexpected results. For example, an open reading frame, rbcX, was juxtaposed between rbcL and rbcS, as in Anabaena sp. strain PCC 7120 (93).

This open reading frame is preceded by a potential ribosome binding site (AGAAGG) starting at a position 18 bp upstream of the putative start codon. Compared with the nucleotide sequence of the Anabaena sp. strain PCC 7120 rbc cluster (120), the space between the rbcL and rbcS structural genes in strain CA (548 bp, including the stop codon[s] of rbcL ) is 3 bp shorter; however, this 3-bp shrinkage occurs outside the coding region of rbcX from Anabaena sp. strain CA (not shown).

Downstream from rbcS, a structure composed of a 13 bp stem and tetraloop, with an estimated free energy of formation of -12.7 kcal (1 kcal = 4.184 kJ)/mol, is separated by 31 bases from the stop codon of the Anabaena sp. strain CA rbcS gene. This structure is tailed with a string of U's followed by an A and might signal a Rho-independent termination signal of transcription (Fig. 3.3A). Analysis of the sequence reported earlier

(120) indicated that a similar stem-loop structure is found at a corresponding position 3' Fig. 3.1. Genomic organization of C02 fixation genes in Anabaena strains. C02 fixation genes are mapped in Anabaena sp. strain CA (A), A, variabilis (B), and Anabaena sp. strain PCC 7120 (C). The locations of the genes, shown by boxes, were determined by DNA sequencing (A), partial DNA sequencing and Southern blot analysis (B), and Southern blot analysis (C). The arrows containing a solid line indicate transcriptional unit and orientation, while arrows with a dashed line indicate only the transcriptional direction of a putative open reading frame. The bars under the Anabaena sp. strain CA map symbolize the fragments used for hybridization probes in this study. C —Clal; D=Dral (the Dral sites are marked only in the rightmost Hindlll fragment of the Anabaena sp. strain CA map); E=£coRI; Ev=£coRV; H=//mdIII; Hp-Hpal, Plasmids 21C1 (75) and pAN600 (31), from A. variabilis and Anabaena sp. strain PCC 7120, respectively, were used to map the rca gene from these organisms.

48 rbcL X S orf 12 rca H HpEvHpH HEHDHpEvD DHpDE H I 11 I *1

1 kb rbcL XS rca E

555555511 ikm //■ IQQQQOjy t _ M

rbcL XS rca region H HHH Hp H

//- «ggs&nH 1J I 1 Fig. 3.2. Nucleotide sequence and deduced amino acid sequences of the region between rbcL and rca in Anabaena sp. strain CA. Nucleotides arc numbered from the stop codon of the rbcL gene to the start codon of the rca gene. The nucleotide sequence encodes 4 open reading frames: RbcX, RbcS, ORF1, and ORF2. Stop codons arc marked by asterisks, and gaps introduced are indicated as dashes. A possible ribosome binding site for the rbcX gene is underlined. The deduced amino acid sequence of RbcX of Anabaena sp. strain PCC 7120 (93) and the RubisCO small subunit sequences of Anabaena sp. strain PCC 7120 (120) and Synechococcus sp. strain PCC 6301 (165) are compared with the RbcX and small subunit sequences of Anabaena sp. strain CA, respectively. The RbcX sequence of Anabaena sp. strain CA is also compared with an open reading frame encoded between the rbcL and rbcS genes of Synechocystis sp. strain PCC 6803 (2), and two frames of amino acid sequences deduced from the nucleotide sequence of the corresponding region in Synechococcus sp. strain a-1 (205) (fl and f2). Residues from the other cyanobacterial strains that are identical to residues found in Anabaena sp. strain CA are not shown.

50 51

TOATAAAQaCTOAAOOTOTAAaaATOAAOOTTAAATAAAAATTACTTCATCCTOAAACAC 60 CA RbcL • * QATACACTTCATCCTTTAQAAQQQCTQQQTCAAGCATGAACCTCAAOCAAATTQCQAAAQ 120 CA RbcX MHLKQXAKD 7120 6803 Q T H Q A a-1 £1 DCQAHCQA a-1 £2 H I V H Q ATACAaCCAAAACACTCCAAAOCTACCTaACTTATCAaaCACTAAaOACTOTOCTOaCAC 180 CA TAKTLQ8YLTYQALRTVLAQ 7120 H 6803 V V V...... a-1 £1 HHQNPD*LSHLSGRAHRDWA a-1 £2 T I A X a AGCTAOGCOAAACAAATCCACCGTTGGCACTTTGGCTGCATAACTTT-- -TCTGCCGQGAAAG 240 CA LGBTNPPLALMLBHF - 8 A G X V 7120 M T - V 6803 a-1 £1 T RNRSTRB Q T Q B 8 a-1 £2 A D RDRSGYTSS - R V K TCCAGGATGGCGAAAAATACATCGAAGAACTCTTTCTCGAAAAGCCTGACTTAGCATTGC 300 CA QD08KYXB8LFL8XPDLALR 7120 A V K R Q 6803 a-1 £1 R L A R Q or a-1 £2 8KHVNATHX 8FANSPXSVF GAATAATGACAGTCAGAGAACATATAOCTGAAGAAATCGCTGAATTCCTACCAGAAATGG 360 CA XKTVRBHXASBXABFIiPBMV 7120 V 6803 LA D 8VLD O T a-1 £1 L L H V D L L a-1 £2 AFSRYAN LRKHWRTXSRKC TTGTGACTGaTATTCAaCAAGCCAACATGGAAAAACGCCGCCAGCATTTAGAACGCATaA 420 CA VTGXQQANKBKRRQBL8RNT 7120 R 8 Q V 6803 RNSLABS X A H BL L a-1 £1 R A L Q Q C Q a-1 £2 YGQASSRPTCNNAVNNB8G* CGCAGGTGAGTTTATCCCATCCCAGTCCTGAGTCAGAACAACAGCAATTTTCCGATCCTG 480 CA QVSLSBP8P8S8QQQF8DPD 7120 B L H T I T 6803 RTVASVDNFP T - - - NOB a-1 £1 BAHVBHBHL T P B *

ACTGGQATAATTTAGCCAGTTAGGAAAATTCCAAGTCGTCACACAATAGCAAACCGTTAT 540 CA H D N L A 8 * 7120 B * 6803 S B N D FPPS*

TATTAQCTATGCAAACCCTACCAAAAGAGCQTCOCTACOAAACCCTTTCTTACCTACCCC 600 CA RbcS NQTLPKBRRYBTL8YLPP 7120 6301 H 8 K F F CCCTCACCGACGCTCAAATTGAGAAGCAAGTTCAGTACATTCTGAATCAAGGCTACATTC 660 CA LTDAQXBKQVQYX LNQGYX P 7120 V 8 6301 8 R AA XB HIE FB

Fig. 3.2 52 Fig. 3.2 (continued)

CAGCAATTGAGTTCAACOAAACCTCTGAACCCACCGAATTTTACTOGACAATGTGGAAOC 720 CA AIBFHBTBBPTBFYWTMWKL 7120 V V L L £301 L I B N B F N

TaCCTTTaTTCAATaCTCAAAaCACCCOCOAAOTATTOaOCaAAaTTCAAOCTTOCCOTT 780 CA FLFNAQBTRBVLOBVQACRS 7120 G R T S A S 6301 DCKBPQQ D RB CTCAATATCCTACCCACTACATCCGTOTTGTAGGTTTCGACAACATCAAGCAQTQCCAAA 840 CA QYPTBYXRVVGFDHIKQCQX 7120 a 6301 B a D C A T TCCTCAOCTTCATCOTACACAAACCCAOCAGATACTAAAAGCTOATTTaGTTTAAATAAT 900 CA LSFXVHXPBRY* 7120 6301 V R O TTATCCATCAAAQAQAQOTAQAQTTTTCTATCTCTCTTTTTTTAOCCACACACTATTAQT 960 GATGCTTGOGTTAATTOOTTTAGATAAGAATTGCTTATACTAGATOTAGCAACACTCCTG 1020

TGAAOAATCTaCCAATaACATCATaACATaAGGCAAOTTTTAGCAOTOTaACAACAAGGT 1080 TTAAACTCTACCGAATATCAGACATTTATACAACACTTATCCAAATAGTCTTAAGGCATA 1140 AAATAAGAATATTGACAGGTTATTCAAGGTGAAATTATGACAGCAATTGCTTTAGATGAA 1200 CA ORF1 NTAXALDB OCOTCTCAACATTTACCAOAATTAATTOAAaCAOCACTGOaCOaTaAGGAAATTATCATC 1260 ASQHLPBLXBAALGOBBXXX

ATaCGAGATAATCAACCTaTTaTAAAGTTAACTCCTaTAaAaCCAGTTAAACaCCaCCaC 1320 MRDNQPVVKLTPVB PVKRRR CAACCTGGAAQTGCCAAAGGCTTAATTACAATAGCTGATOACTTTTTTTAGTATTGCCCA 1380 QPGSAKGLXTXADDFF* CCTGACTGTAATTGCAATACAGTAATTGOTGCAAGATGTAAGTAAATGAATATTGTAAAA 1440 TTACTAACXTTTTATTACCATOTTTAATAACTTAOATTTTCTTAAAGCACACCCAGCATT 1500 CA ORF2 HFNNLDFLKAHPAP

CGATAAG CTAGAAATTATTAAAAAAGGAACTTATTATACAATCTATCGOCCAAAGACTGA 1560 DKLEXXK1CGTYYTXYRPKTD CAGTGATAGGGATTGTTACAAAACTTTATTTCTACAATATTTCAAAAACTTTGTTATTCT 1620 SDRDCYKTLFLQYFKNFVXL TATTGAATATGGATGTTTTATTGATCATTTTATAATTATTGAACGAAATTATTTTGACAA 1680 XBYOCFXDHFXXXERNYFDH TACTAAAG CACATAACTTAAAAGAAAAATGTTACCGTTTTTTACCAAAGCTAGATOAGGA 1740 TKAHNLKBKCYRFLPKLDBD TAATAGTAGCCGACCTCAAGATGGAGACCTTTTAGATGAAAAAAAATATTGGTCTATTCT 1800 NSSRPQDGDLLOBKXYWSXL CAATGATAGCTTTGTTCAAGAAATAGAAATATGTTATCGCAGGATTATAGATACAGAAGA I8 6 0 NDSFVQBXBXCYRRXXDTBD 53 Fig. 3.2 (continued)

TACCCATACAAOTAOCTATAOATTAOTATCTOOTOOTATTTTaCATACCAOAaCTATTaC 1920 THTSSYRLVSGGZLHTRAIA TACTTTATTTAATCTTCTGTCTGCOCTTGAAAATATATGGAACTCTTTCAAAOCATTAAA 1900 TLFHLLflALBHZHNSPKALH TATAGTTAAAAOAACAAGATTTTTGCCTATCTCTTCTACQAATTTCCCCCACATTACATC 2040 XVXRTRPLPXS8TNFPHXTS AAAAGACTATTTAGAAOACACAAACATAAATAAATTGTCTQATCTTAAOGAAGTTATCCT 2100 XDYLBDTHXHRLSDLRRVZL TTTOAATGATGACGAAAQTAQTTACCQTTATTTCTATCCTGAAQATGTAOAAQCAGAAAA 2160 LHDDBSBYRYFYPBDVBA8K GGAACGTCGTCGCAAAGAAGCAGCAOCAAOAAAOAAGGAACGGGAGCAGGAAAGAAAGGA 2220 BRRRKBAAARXK8RBQ8RX8 ATATATQCAACTAGAGaCAAaAGAGTaGQTAAATTTTGGATATTATaCAGCTATGOTTGA 2200 YNQL8ARBWVNF0YYAAMVD CaAOOOAOAATTAaCTCCTOATGATCCTAOATATCCCTTCTaATCAaAAaTOCCACTTTA 2340 BGBLAPDDPRYPF*

TCTGGGTATTAAGTTTTTTTTAGTTOAGCTACTGAaCAACCAGGAATTTTTACCAAAAGA 2400 TTAGTAAACATAAQATACCAAAAAAATTTAACATTAATTGAATACAACATTATCAGTGAT 2460 AATTTTATO 2469 CA Re* N 54

- I S . 7 K cal -1 S .C K cal

|rbcs stop] ^Sl nt |ri>cg Step] / j i n t uuuuouuuuua

A nabaena CA Anabaena 7120

- 1 0 .4 K cal u u U A c - a A -u o - c - 1 9 .8 K cal a - c A-u A o - c C A U-A U A U-A c u c - o U-A U-A U-A A-U c - a U-A A-U U-A o -c C-O c -o C-O c -o C-O c -o O-C A-U A-U U-A U-A A-U C-O O-C A-U

Anabaena CA

Fig. 3.3. Potential transcriptional terminators for Anabaena rbcLXS and rca. The stem- loop structures and their free energy of formation were predicted by using the RNAFOLD program of PC/GENE (IntelliGenetics, Inc.). nt, nucleotides. 55 from rbcS in Anabaena sp. strain PCC 7120 (Fig. 3.3B) (120). Interestingly, the 3' noncoding region of the Anabaena sp. strain CA rca gene contains two inverted repeats in the likeness of a Rho-depcndent transcriptional terminator (Fig. 3.3C). A 20-bp repeat starts at a position 28 bases 3' to the stop codon of rca and has a free energy of formation of >30.4 kcal/mol; a second repeat of 13 bp is situated 131 bases downstream from the stop codon of rca and has a free energy of >18.8 kcal/mol.

Since chromosomal DNA from several Anabaena and Nostoc strains was shown to hybridize to Anabaena sp. strain CA and spinach rca sequences (98), plasmids containing the rbcLrbcS genes of A. variabilis (75) and Anabaena sp. strain PCC 7120

(31) were probed with the Anabaena sp. strain CA rca gene. As in Anabaena sp. strain

CA, the rca gene is downstream from rbcS in A. variabilis and PCC strain 7120.

Although theAnabaena sp. strain PCC 7120 rca gene was not precisely mapped, the rca gene appeared to be much more distant from rbcS in this strain than the others (Fig,

3.1). Partial sequencing confirmed that the rbcL, rbcS and rca genes of A. variabilis were also transcribed in the same direction. Situated between rbcS and rca in Anabaena sp. strain CA, there are two unknown open reading frames (ORF1 and ORF2), and all of the genes ( rbcL, rbcS, ORF1, ORF2, and rca) are in the same transcriptional orientation (Fig. 3.1 and 3.2). When the distribution of ORF1 and ORF2 was examined in the otherAnabaena species, or the unicellular and nonheterocystous filamentous strains previously examined (98), in no case was there hybridization to the Anabaena sp. strain

CA ORF1-ORF2 probe under conditions of low stringency. Likewise, there was no indication that ORF1 and ORF2 were found in the intergenic region between rbcS and 56 rca in A. variabilis and Anabaena sp. strain PCC 7120, suggesting that these open reading frames arc unlikely to be a fundamental part of the C 02 fixation regulatory machinery.

Light/dark and nutrient effects on transcription. In previous preliminary studies, for unknown reasons rca transcripts were not detected in Anabaena sp. strain CA under conditions in which rbc transcripts were readily obtained (98). However, with the protocols outlined in this study, rca specific mRNA was readily detected. Furthermore, it was found that rbcL and rca transcripts accumulated in the light in Anabaena sp. strain CA; transcripts were not found in cells kept in the dark for 16 h (Fig. 3.4A, 3.4D, and 3.4C). Once the cells were returned to the light, the rfccL-specific transcripts (2.7,

2.2 and 1.85 kb) and the rat-specific transcript (1.35 kb) rapidly appeared (Fig. 3.4B and 3.4C). As a positive control, we also determined the levels of psbA and ni/H transcripts, since the expression of psbA (117) and the synthesis of nitrogenase (135) in other cyanobacteria were found to be light regulated. Only trace amounts of psbA and niJH transcripts were detected in the dark in Anabaena sp. strain CA (Fig. 3.4B and

3.4C, respectively), and this was probably due to partial initiation of mRNA synthesis caused by unavoidable irradiation during cell harvest, an interpretation consistent with

RNA degradation rates (Fig. 3.5). The maximum levels of the rca and psbA transcripts, however, were reached after 5 min of exposure to the light, while the amounts of rbcL and niJH mRNAs displayed a more gradual time-dependent increase (Fig. 3.4B and

3.4C). Although the results were not quantitated, cells grown at 1% and air-levels of

C02 showed parallel accumulation patterns of rbcL and rca transcripts. This is unlike Fig. 3.4. Northern blot hybridization analysis of gene expression in Anabaena sp. strain CA. (A) RNA was isolated from cultures grown in NH*N03-containing medium, bubbled with 1% C02 in air (lanes 1 and 3) or air (lanes 2 and 4), under constant illumination (lanes 1 and 2) or transferred to the dark for 16 h before harvest (lanes 3 and 4). (B) RNA was isolated from cultures grown in combined nitrogen-free medium, bubbled with air, with 16 h of dark treatment as described for panel A (lane 1) or shifted back to the light after the dark treatment for 5 min (lane 2), 10 min (lane 3), or 20 min (lane 4). (C) Same as panel B except that the cells were bubbled with 1% C02 in air. (D) Cells were grown in NH4N03-containing medium, under constant illumination, bubbled with 1 % C02 in air (lane 1) and subsequently bubbled with argon for 4 h with cells collected at 1 h intervals (lanes 2 to 5). For the 3-h time point (lane 4), considerably less than the normal 15 fig of RNA was applied. The probes used are indicated above the blots. Sizes (in kilobases) of hybridized mRNA species are indicated.

57 58 » I * I

1 * § * '

B rki. 12 1 4 1114 I f} .

zr 22 V/» 185

135 12

rbcL nifll 12 14 12 14

M 095

rfcrj ( M rra 12)41 12)41 12)4)

Fig. 3.4 59 the situation in Rhodobacter sphaeroides (58), in which case cells exposed to excess carbon exhibited a repression of RubisCO gene expression. A slight decrease in rca mRNA from 0 to 1 h was not accompanied by substantial changes in the levels of rbcL, rbcS, and rca transcripts after 4 h of carbon starvation of Anabaena sp. strain CA (Fig.

3.4D), making it unlikely that exogenous carbon quantitatively affects transcription of these genes.

The source of nitrogen is a key determinant in regulating hetcrocyst differentiation in Anabaena species, and the addition of combined (fixed) nitrogen can totally inhibit the expression of genes specific for nitrogen fixation (66). On the other hand, the levels of rbc transcripts remain constant during the induction of heterocyst differentiation, i.e., when ammonia-grown cells are transferred to a nitrogen-free medium (92). Since transcription of the rbcL and rca genes appeared to be differentially controlled, we examined the effects of the nitrogen source on the light regulated expression of the rbc and rca genes. Light-induced accumulation of rbcL and rca mRNAs was seen in both

NH4NO,-utilizing cultures and Nyfixing cultures (Fig. 3.4A, 3.4B and 3.4C), indicating that the source of nitrogen had no effect on the pattern of rbcL and rca mRNA accumulation in the light or the dark. However, when we attempted to determine if the transcription of ORF1 and ORF2 was similarly controlled, we were unable to detect transcripts in Anabaena sp. strain CA that hybridized to the ORF1/ORF2 DNA probe.

To further characterize the influence of light on transcription, the decay of rbcL, rca, psbA, and nifH transcripts was monitored both in the light and in the dark (Fig.

3.5). The levels of rbcL, psbA, and niJH transcripts declined slightly more rapidly in the c 1 00 S O E (0

8> 60-

01 E 0) £ 3 & 140 Time (min)

(B) 120 Dark V 100

at O) at E

I 3 & 140

Time (min)

Fig. 3.5. Stability analysis of Anabaena sp, strain CA mRNA. Decay of transcripts was measured in the light (A) and in the dark (B) following the addition of rifampcin. RNA was extracted 0,15,30,60, and 120 min after treatment. The relative amounts of transcripts for each indicated gene are plotted as a function of time. light than in the dark. Ha!f*Hves of rbcL, psbA, and niJH transcripts were estimated to be 19.4, 33.1, and 16.2 min in the light and 31.2, 46.0, and 27.2 min in the dark, respectively. The turnover of rca messages proceeded much faster than that of rbcL transcripts. Only trace amounts of rca transcripts remained 15 min after the first addition of rifampcin under both illumination conditions; the half-life of rca transcripts was not longer than 7.1 min in the light and 7.9 min in the dark. Since A. variabilis is able to utilize fructose as a carbon source both in the light and in the dark (73), we considered the possibility that fructose modulates the transcription of C02 fixation genes in this organism. Photohetcrotrophic cultures of A. variabilis accumulated rbcL and rca transcripts to levels comparable to those in photoautotrophic cultures; however, 10 mM fructose partially released the dark repression of rbc and rca transcription, such that there was about 20% expression in the dark compared with cultures maintained in the light

(Fig. 3.6).

The 5' end of the rca transcript was mapped after isolating mRNA from N2-fixing

Anabaena sp. strain CA cultures bubbled with 1 % C02 in air under constant illumination.

Three contiguous initiation sites were detected at positions -27 to -25 with respect to the start of translation (Fig. 3.7A). Under these growth conditions, the RNA initiating at the -25 position was the most abundant rca message species. The Anabaena sp. strain

CA coding sequence, not including the stop codon, is 1,245 bp; combined with the fact that the message begins at positions -27 to -25 and presumably ends after the first repeat in the 3' noncoding region, the transcript should be approximately 1,350 bp, precisely what is observed (Fig. 3.4). Finally, the 1.7-kb EcoRV-Hpal rca-encoding fragment 62

fbcL rca 1 2 3 4 1 2 3 4

Fig. 3.6. Influence of fructose metabolism on rbcL and rca transcription in A. variabilis. RNA was isolated from cultures grown in the absence (lanes 1 and 2) and presence (lanes 3 and 4) of fructose, In lanes 1 and 3, cultures were incubated under constant illumination, while the cultures in lanes 2 and 4 were shifted to the dark for 16 and 73 h, respectively. Sizes (in kilobases) of hybridized mRNA species are indicated. Fig. 3.7. Primer extension analysis of the 5' end of the Anabaena sp. strain CA rca mRNA and role of the 154 bp 5' noncoding sequence in recombinant RubisCO activase synthesis. (A) Results of the primer extension reaction, Nucleotides antisense to the reaction products are indicated on the left. (B) Summary of results. Arrows indicate the transcription start sites of rca\ the shaded sequences are plausible -35 and -10 regions, compared with a consensus E. coli sequence. A hexamer with a sequence identical to the putative -35 region of the Anabaena sp. strain PCC 7120 psbA promoter (48) is underlined. (C) The role of the 154-bp upstream noncoding sequence in supporting recombinant RubisCO activase synthesis was evaluated by examining E, coli JM109 strains containing plasmids pK18 (lane 2), pK18::EV1.7-l (lane 3), and pK18::EV1.7-2 (lanes 4 and 5). 0.5 mM IPTG was added to some cultures (lanes 2 and 5) when the cells reached an A ^ of 0.5. Purified recombinant RubisCO activase (lane 6), verified by Western immunoblot analysis and N-terminal sequencing analysis (98) was used as a standard. Lane 1, commercial molecular weight standards (from top to bottom, with molecular weights of 97,400, 66,200, 45,000, 31,000, and 21,500). LacZa, a-peptide of h-galactosidase.

63 64

(A) t C O T

(B) -33 -10 ill - _ Tf? A.CA S'-TlMHI^AMlATWXMUUMAATlfl^BlMTraUTACAACATrAT JT. c o l i TTOACA TATAAT

CAaTOATAATTTTATO 70 n t — — CTOOTOrOCOOOTTCCCTTO-3 ' j ^ p r i m e r ~j

Fig. 3.7 65

Fig. 3.7 (continued)

< ■ Rca

L a c Z o t^ 66 (Fig. 3.1A) which contains 154 bp of 5' noncoding sequence was cloned in both orientations with respect to the lac promoter of plasmid pK18. Recombinant RubisCO activase was synthesized in E. coli in either orientation without a substantial requirement for isopropylthiogalactoside (IPTG) induction (Fig. 3.7C). These results indicated that the 154-bp upstream sequence contained a region that was recognized as a promoter by the E, coli host.

DISCUSSION

Doth rbcL and rbcS hybridized to transcripts of 2.7 and 2.2 kb when Northern blots were used to analyze specific mRNA isolated from Anabaena sp. strain CA (Fig.

3.4D), indicating that the rbcL and rbcS genes of Anabaena sp. strain CA are cotranscribed as in other prokaryotes (58,60,89,112,120,166,172,202). Since rbcX is located between rbcL and rbcS, it is undoubtedly part of the rbc operon. On the other hand, many of our data point to the fact that the rca gene may not be part of the rbc operon in Anabaena species. For example, rbc and rca probes hybridized to distinct transcripts in both Anabaena sp. strain CA and A, variabilis, and Anabaena sp. strain

CA and Anabaena sp. strain PCC 7120 both contain potential transcriptional terminators immediately following rbcS (Fig, 3.3A and 3.3B). In Anabaena sp. strain PCC 7120, therbcLrbcS and rca genes are fairly distant, making it unlikely that these genes are part of the same transcriptional unit; also the size of the Anabaena sp. strain CA and A. variabilis rca transcripts suggests that the rca mRNA may be monocistronic. Finally, the primer extension experiments, in which the presumptive 5* end of the Anabaena sp. 67 strain CA rca transcript was mapped, and the discovery of transcription terminator-like structures downstream of the Anabaena sp. strain CA rca gene are both compatible with the suggestion that the Anabaena rca gene forms an operon of only a single gene. In other bacteria that assimilate carbon dioxide, such as R. sphaeroides (57, 58) and chemoautotrophic bacteria (112, 202), genetic studies led to the unequivocal conclusion that the Calvin cycle structural genes are all part of a single large rcgulon containing several genes (56). In these organisms, a large operonic transcript is cleaved, apparently in a specific fashion, to yield several small transcripts. Thus, it will be important in future genetic experiments to determine if a similar scenario is true for Anabaena, since specific endonucleolytic cleavages of a single large transcript containing the rbc and rca genes could account for the results thus far obtained.

Since RubisCO is synthesized only in vegetative cells of free living heterocystous cyanobacteria (28, 65), one might expect that rca is expressed only in vegetative cells; this, however, must be established experimentally. Despite the fact that no specific consensus sequence has been defined for promoters from Anabaena species, the primary

RNA polymerase isolated from Anabaena sp. strain PCC 7120 vegetative cells seems to prefer E. co/Mike promoters (157). Hexamers TTGACA and TATAAT, with a gap of

17±1 bases between each other and a space of 6±1 bases between the latter hexamer and the start of transcription, represent the consensus -35 and -10 regions of E. coli promoters (71). Thus, the shaded sequences, TAAACA and TAACAT (Fig. 3.7B), are plausible -35 and -10 regions, respectively, of the Anabaena CA rca promoter. The sequence, TAGTAA, overlapping with the TAAACA sequence (Fig. 3.7B), could be 68 another candidate for the *35 region of the Anabaena sp. strain CA rca promoter, since

TAGTAA has been proposed to be the -35 clement of the Anabaena sp. strain PCC 7120 psbA promoter (157). The expression data (Fig. 3.7C) indicate that a promoter sequence is present on the 154-bp upstream sequence, which includes the homologous -35 and -10 regions. Although these results do not directly demonstrate the existence of an Anabaena rca promoter, in combination with the primer extension studies and the sequences noted above, these findings suggest that this region should be the focus for future studies.

RubisCO not only uses C02 as the gaseous substrate for carboxylation but also requires another molecule of C 02 for activation (104). One of the functions of plant

RubisCO activase is to somehow lower the C 02 concentration required for activation so that RubisCO activity can occur at physiological levels of C 02 (126). Therefore, it was assumed that the concentration of C02 supplied to Anabaena cells would be critical for the maximum expression of rbcL, rbcS and rca. However, Northern blot hybridization studies did not support this hypothesis, since the levels of rbc transcripts were indistinguishable, and the levels of rca transcripts were very similar when a culture was switched from 1 % C 02 in air to an argon atmosphere. These findings may be related to the fact that some cyanobacteria have a powerful C02-concentrating mechanism which actively assimilates inorganic carbon (Q) and subsequently raises the concentration of

C02 in the carboxysomes, polyhedron-like structures which contain most of the intracellular RubisCO of cyanobacteria (48, 131). Since bicarbonate transport may be induced by limiting C, (131), the intracellular C02 concentration may not fall with decreasing external Ct in cyanobacteria. This may be the reason why transcription of 69 rbcL and rca was maintained at a level which was apparently independent of the external

C02 concentration. Alternatively, the effects of C02 may be exerted at some level after transcription, as in Chlamydomotws reinhardtii. This unicellular green alga also possesses an inducible C02-concentrating mechanism and exhibits no quantitative differences in rbcL and rbcS mRNAs at either high levels of exogenous C 02 or limiting

C02 levels (201). However, induction of the C02-concentrating mechanism by limiting

C02 was accompanied by a decrease in the synthesis of RubisCO subunits in C. reinhardtii. Thus, it was suggested that the decline of RubisCO synthesis in this organism might be caused partly, if not totally, by modification of the rbcL and rbcS transcripts.

Although not absolutely proven by our experiments to date, all of our data are consistent with the conclusion that the accumulation of rbc and rca transcripts in

Anabaena species is regulated by light as in plants (152, 168). For the strict autotroph

Anabaena sp. strain CA, the accumulation is absolutely light dependent, and light control appears to be exerted at the level of transcription rather than message stability.

The half-lives of theAnabaena sp. strain CA rbcL and rca transcripts in the dark are approximately 1.6- and 1.1-fold, respectively, greater than those found in the light. In the heterotrophA. variabilis , fructose metabolism leads to a relaxation of the dark control of transcription, since transcription of rbc and rca is dimished but not completely turned off in the dark during heterotrophic growth. In other words, some degree of light dependence still exists in fructose-utilizing cultures of Anabaena variabilis. In contrast, in the symbiont Anabaena azollae, the levels of rbc transcription were similar in the light 70 and in the dark (120). Since RubisCO activity has been suggested to be important for the dissipation of reducing equivalents in pholoheterotrophic cultures of Synechocystis sp. strain PCC 6803 (123) and R. sphaeroides (193), RubisCO may serve the same function when A. variabilis is grown on fructose.

Other findings might relate to the physiological' role of RubisCO activase and

RubisCO in Anabaena species. For example, it was observed that the level of rca mRNA fluctuates instantly in dark-light and light-dark transitions of Anabaena sp. strain

CA, whereas rbcL transcripts accumulate and fade in a progressive manner (Fig. 3.4B,

3.4C and 3.5B). Since the presumed role of RubisCO activase is to influence the activity of RubisCO, a sharp rise or fall of the level of RubisCO activase might lead to a great change in the activation level of RubisCO. In this scenario, activity could be tuned up or down in spite of the total amount of RubisCO protein present. Thus, the more rapid adjustment of rca transcription expression compared with rbc transcription as a result of changes in illumination is consistent with the proposed regulatory role of activase in modulating RubisCO activity. Obviously, further studies should be directed at the level of RubisCO activase and RubisCO protein and enzymatic activity under these growth conditions; the transcription studies reported here should provide a useful framework for the subsequent design of such experiments.

There are considerable differences in the deduced amino acid sequences of

Anabaena and Synechococcus RubisCO small subunits (Fig. 3.2). Although the binding sites for activator C02 and MgI+ are located on RubisCO large subunits (103) and the small subunit has no effect on inactivation caused by RuBP (95), the small subunit does 71 have an impact on the inhibition mediated by other phosphorylated metabolites, e.g., 6* phosphogluconatc (95,137) and the transition state analog (169). Perhaps the divergence in the primary structure of small subunits leads to a differential requirement for RubisCO activase, since Synechococcus DNA yields no clear hybridization signal towards an

Anabaena sp. strain CA rca probe (98), although some form of activase protein is reportedly present in Synechococcus sp. strain PCC 7942 (48).

Finally, in both Anabaena sp. strain PCC 7120 (93) and Anabaena sp. strain CA

(this study), the rbcX gene is located between rbcL and rbcS, and it encodes a polypeptide with a predicted Mr of about 15,000. The RbcX protein of Anabaena sp. strain CA contains an additional three residues at the C terminus and shares 84% identity to the RbcX protein of Anabaena sp. strain PCC 7120 (93) (Fig. 3.2). The rbcL-rbcS intergenic region of Synechocystis sp. strain PCC 6803 (2) also accommodates an open reading frame; its putative gene product (Mr of approximately 10,000) displays 45% identity to the RbcX protein of Anabaena sp. strain CA when a large gap is introduced between amino acid residues 23 and 24 (Fig. 3.2). Interestingly, one frame of amino acid sequence deduced from the nucleotide sequence between rbcL and rbcS genes of the thermophile Synechococcus sp. strain a-1 (205) exhibits significant homology to the C terminus of the Anabaena sp. strain CA RbcX protein and contains an ATG triplet at a position corresponding to the start of the RbcX protein, while the +1 frame shows similarity to the N terminus (Fig. 3.2). The situation is quite different in Synechococcus sp. strain PCC 6301, since the intergenic region is only 93 bp (165). The deduced amino acid sequence of rbcS from Anabaena sp. strain CA shows 88 and 67% identity 72 to small subunits of Anabaena sp. strain PCC 7120 (120) and Synechococcus sp. strain

PCC 6301 (165), respectively. Although the arrangement of the rbc operon is similar in Anabaena strains, ORF1 and ORF2, situated between rbcS and rca in Anabaena sp. strain CA, are absent in the corresponding intergenic regions in A. variabilis and

Anabaena sp. strain PCC 7120 (Fig. 3.1). In addition, culture conditions tested in this study failed to yield detectable expression of the two unknown open reading frames in

Anabaena sp. strain CA. The identities of rbcX , ORF1, and ORF2 remain mysterious, but their close association to structural genes critical for C02 fixation is intriguing.

Moreover, since we are interested in factors that regulate C02 fixation in oceanic cyanobacteria, the conditions used to delect active transcripts with the coastal marine organism Anabaena sp. strain CA should provide the basis for initiating gene regulation studies with oceanic strains, both in the laboratory and in the field. CHAPTER IV

Regulation of Anabaena RubisCO in vitro and in a RubisCO Activase Mutant

INTRODUCTION

Like most autotrophic organisms, cyanobacteria fix C02 into organic carbon primarily by the reaction catalyzed by ribulose bisphosphate carboxylase/oxygenase

(RubisCO). Cyanobacterial RubisCO is a multimeric enzyme composed of eight large subunits and eight small subunits. The genes encoding the large and small subunits (jbcL and rbcS , respectively) are closely associated and cotranscribed (99, 120, 166). The cyanobacterial rbc operon has been successfully expressed under the control of a lac promoter in Escherichia coli (70, 185). However, the yield of active recombinant

Anabaena 7120 RubisCO is low, presumably due to poor expression of small subunit polypeptides. It was hypothesized that secondary structure within the intergenic region between the Anabaena 7120 rbcL and rbcS genes might be recognized as a transcription terminator at high frequency by £. coli, with the result that this early termination event leads to low production of small subunits (70). Assembly of the LgSg holoenzyme in chloroplasts requires a protein, the RubisCO-binding protein (150), which shares 46% identity with the "molecular chaperone", chaperonin 60 (cpn60 or GroEL) of E. coli

(74). Overexpression of cpn60 and its companion cpnlO (GroES) enhances proper

73 74 assembly of cyanobacteria! RubisCO in E. coli (68, 93).

RubisCO is catalytically quiescent until activated by C 02 and Mg1*. The activator C02 covalently binds to the c-amino group of a lysine residue (Lys-201) of the large subunit at the active site to form a carbamate, which enables the enzyme to associate with the catalytically essential Mg1* ions (103). In plants, the level of

RubisCO activation increases when shifted from dark to saturating illumination and it was shown that light activation requires an enzyme called RubisCO activase (153, 154); however, the mechanism by which activase promotes RubisCO activation is not well understood. Several studies also link RubisCO activase to alleviating the inhibition of

RubisCO activity caused by the binding of sugar metabolites at its active site (for a review, see ref. 127).

Ribulose 1,5-bisphosphale (RuBP), the substrate for RubisCO, binds tightly to the noncarbamylated form (E form) of the enzyme. The ER form of the enzyme obstructs the formation of the active carbamylated metal-bound enzyme complex (ECM form) (81).

RubisCO activity also progressively declines after the addition of RuBP to the activated

ECM ternary complex (38). This time-dependent "failover" in RubisCO activity is the result of the generation of inhibitory compounds via the isomerization of RuBP into 3- ketoarabinitol 1,5-bisphosphate (KABP) and xylulose 1,5-bisphosphate (XuBP) during catalysis (Fig. 4.1) (41,209). KABP exhibits a high affinity for the ECM form whereas

XuBP prefers the E form (39, 40, 208, 209). The relative amounts of the two phosphorylated sugars formed during catalysis and their affinity for RubisCO varies with pH, although isomerization occurs at a relatively constant rate. KABP is the Fig. 4.1. Isomerization of RuBP during carboxylation. During RubisCO-catalyzed carboxylation, ribulose 1,5-bisphosphate (RuBP) is converted intoanenediol intermediate before the addition of C02 and subsequent split into two molecules of 3-phosphoglyceric acid (PGA). The enediol intermediate can regenerate RuBP via protonation. Misprotonation at C-3 or C-2 leads to the formation of xylulose 1,5-bisphosphate (XuBP) and 3-ketoarabinitol 1,5-bisphosphate (KABP). [Modified from Edmondson et al., 1990 (41)].

75 I 41 HOCH I KABP C=0 HCOH I c h 2 o p

ch2° p CH2° P CH2° p c=0 "OOC-COH 1 C02 | ch2° p HCOH COH ► c=o HCOH I I I HCOH HCOH HCOH COOH I I I ch2 o p ch o p 2 ch2 o p PGA RuBP e n e d i o l

!«•

OJ2OP c=o I XuBP HOCH I HCOH I Hg. 4.1 predominant byproduct and causes failover alone at pH 8.5; at pH 7.5, XuBP exceeds

KABP and contributes about 40% to failover (209). Most plant species tend to accumulate another phosphorylated compound, 2-carboxyarabinitol 1-phosphate (CA1P), in the dark, but to different extents. Because CA1P structurally resembles the transition- state intermediate of the carboxylation reaction, it readily binds to carbamylatcd catalytic sites, blocking carboxylation. Thus, variation of the intracellular CA1P pool under light and dark conditions accounts for part of the light regulation of RubisCO activity (for review, see ref. 162). RubisCO activase has been shown to facilitate the dissociation of phosphorylated inhibitors, including RuBP, KABP, XuBP, and CA1P, from the active sites in an ATP-dependent manner, thus increasing the availability of activated RubisCO and allowing activation to proceed (143, 144, 194).

The RubisCO activase gene (rca) is present in hcterocystous cyanobacteria such as Anabaena and Nostoc strains (98). In three strains of Anabaena, the rca gene is closely associated and downstream from the rbc operon. In Anabaena sp. strain CA, the accumulation of RubisCO and RubisCO activase transcripts is differentially regulated by light (99). This report further investigates the functions of RubisCO activase in

Anabaena spp. Recombinant Anabaena sp. strain CA RubisCO and RubisCO activase were purified from extracts of E. coli for in vitro studies. It was found that recombinant

Anabaena RubisCO was not inhibited, or inhibited slightly, by known sugar metabolite inhibitors. In addition, a rca mutant of Anabaena variabilis was constructed by target mutagenesis. This mutant exhibited a preferential "high C02-requiring" phenotype and was unable to elevate RubisCO activity to a level shown by wild type when cultured 78 under high light intensity in a 1% C02-air atmosphere.

MATERIALS AND METHODS

Media and growth conditions. A. variabilis sp. strain 29413FD and its mutant derivative were grown in BG-11 media (141) at 28° C, sparged with air or 1% C02 in air, under fluorecent lights at 8 (low light), 44 (medium light) or 73 (high light) /imoles quanta m‘z s'1. For solid media, 1.5% agar and 0.3% Na2S202 were added to the BG-11 growth media. To select a mutant containing an insertion of a streptomycin/spectinomycin resistance (SmVSp1) cartridge, these two antibiotics were added to the media at 2.5 /ug/ml each. E. coli was cultured as previously described (98).

The following antibiotics were supplemented as needed: ampicillin, 100 fig/ml; chloramphenicol, 30 pg/ml; kanamycin, 50 /xg/ml; streptomycin, 50 pg/ml; spectinomycin, 50 /ig/ml; tetracycline, 25 ng/ml; trimethoprim, 200 /zg/ml.

Construction of plasmids for recombinant protein expression. Plasmid pLl has the EcoRV-EcoRI rfaL-cncoding fragment of Anabaena sp. strain CA directly cloned into the Hincll-EcoKl sites of plasmid pK19 (134). Plasmid pS5, containing the EcoRI-

Dral rbcS-e ncoding fragment, was subcloned from plasmid pRK415::E4.3, which contains the 4.3-kb EcoRI rteS/rat-encoding fragment in pRK415 (84). This was accomplished after double digestion at BamHl, a site within the multiple cloning region of the vector, and Dral, followed by ligation to the ZtowiHI-S’mal sites of pK19. The

Pvull fragment containing the P lac~rbcS fusion of pS5 was subsequently ligated into the 79 Nhel site of pLl, yielding plasmid pLS23. Plasmid pLS32 contains the EcoRV-Dral fragment, which encodes the whole rbcLXS operon, behind a single lac promoter (Fig.

4.2). The construction of the RubisCO activase expression plasmid pK18::EV1.7-2 was described previously (Fig. 4.2) (98).

Expression and purification of recombinant proteins. Induction of recombinant protein synthesis in E. coli and the preparation of cell extracts were exactly as previously described (136). Recombinant RubisCO was purified through Green A-agarosc column chromatography, ammonium sulfate precipitation, sucrose density gradient centrifugation, and Mono Q column chromatography using a buffer containing 25 mM Tris-HCI, pH

8.0, 1 mM EDTA, 5 mM fl-mecaptocthanol, 10 mM MgCl2 and 50 mM NaHCOj essentially as described by Read and Tabita (136) with slight modification. Ammonium sulfate was not removed by dialysis; instead, proteins were precipitated again by adding two fifths volume of 50% polyelhelene glycol 8000 to the dissolved ammonium sulfate precipitate. In addition, a 0.2-0.8 M sucrose linear gradient rather than a step gradient was employed. The procedures for purification of recombinant RubisCO activase were described in Chapter II (98). Like RubisCO, dialysis was eliminated and polyelhelene glycol precipitation was employed prior to Q-scpharose chromatography. Purification was monitored by SDS polyacrylamide gel electrophoresis (105).

Enzyme assays. RubisCO activity was determined by the RuBP-dependent incorporation of ,4C02 into acid-stable 3-phosphoglyceric acid (199). Carboxylation was assayed at 30° C in HEPES, pH 8.0, plus 20 mM NaHCOj (containing 8 pCi of

NaH,4COj), 10 mM MgCI2 and 0.8 mM RuBP. Usually RuBP was added to initiate 80 carboxylation 5 min after the addition of NaHCOj and MgCI2 (two-step assay); a one-step assay was performed by adding all the reagents simultaneously. Reactions were terminated by the addition of propionic acid. The recombinant Anabaena RubisCO was purified in the ECM form and was directly used to examine the inhibitoiy effect of CA1P preincubation. To examine the effect of RuBP, the E form of RubisCO was generated by gel filtration of the purified enzyme through an Econo-Pac I0DG column (Bio-Rad) using a 0.1 M HEPES buffer, pH 8.0. RuBP was then added to the E form of RubisCO at a level of 0.5 /imolcs RuBP per mg protein per ml. The RuBP-bound unactivatcd enzyme (ER) was then treated with 20 mM NaHC03, 10 mM MgCI2 and 4 mM RuBP for various time intervals (activation step) before carboxylase activity was measured in a 1-min assay by adding the activated enzyme to the RubisCO assay solution described above (one-step assay) (146). For the whole-cell RubisCO assay, wild type and mutant cultures of A. variabilis , with an A^, ranging from 0.3 to 0.5, were washed and subjected to toluene treament as described previously (182). The toluene-treated cells were washed twice with 0.1 M HEPES, pH 8.0, prior to performing a one-step RubisCO assay. The ATPase activity of purified RubisCO activase was determined by measuring the rate of NADH oxidization in a coupled spectrophotometric assay using pyruvate kinase and lactate dehydrogenase (145). Reactions were initiated by adding RubisCO activase to assay mixtures containing 50 mM Tris-HCl (pH 8.0), 10 mM MgCI2, 1 mM

ATP, 1 mM phosphoenolpyruvate, 0.3 mM NADH, 40 units/ml pyruvate kinase and 40 units/ml lactate dehydrogenase, Protein concentration was determined using the Bio-Rad protein assay reagent according to the protocol supplied by the manufacturer. The 81 protein content of whole cells was measured by a modified Lowry protein assay method

(173) after cells were boiled in 2 N NaOH for 30 min (12).

Targeted mutagenesis of the A. variabilis rca gene. A 3.7-kb EcoRI fragment containing the A. variabilis rca gene was subcloncd from plasmid 21C1 (75) into pK19

(134) to generate plasmid pVA64K. The region between two adjacent Clal sites within therca gene was subsequently displaced with a 2.0-kb Sm7Spr cartridge, obtained from

Smal digestion of plasmid pHP450 (129). The EcoRI fragment containing the mutated rca gene was then cloned into plasmid pNOT19 (159) to form pVA64N. A Notl fragment containing oriT, sacRB and the trimethoprim resistance marker (Tmr) from pMT289 (a derivative of pMOB3 (159) generated via a displacement of the chloramphenicol resistance marker with Tmr in a BamUl fragment) was then inserted into the single Notl site of pVA64N, thus generating plasmid pVA353 (Fig. 4.5A). Plasmid pVA353 was cotransformed with helper plasmid pRL528 (42) into E. coli JM109. A triparental conjugation was performed with A. variabilis, E. coli JM109(pVA353 & pRL528) and E. coli ED8654(pRL443) according to the method of Elhai and Wolk (42).

The Smr/Spr Anabaena transformants were examined by Southern blot hybridizations following the protocol supplied with GeneScreen Plus membranes.

RESULTS

Recombinant RubisCO synthesis. No typical stem-loop transcriptional terminator structures, except perhaps for sequences that appear to show relatively 82 insignificant folding capacity, were found within the intergenic region between the rbcL and rbcS genes of Anabaena sp. strain CA by computer analysis. Thus, these two genes were seperately cloned for heterologous expression. Since the region which might lead to RNA folding is in close proximity to the 3' end of rbcL , the £coRI site situated in the center of the intergenic region within rbcX is a convenient site to detach the rbcL and rbcS genes. The /TcoRV-EcoRI rhcL-encoding fragment and the EcoVH-Dra\ rbcS- encoding fragment were ligated behind the lac promoter of pK19 to form plasmids pLl and pSS, respectively (Fig. 4.2). The p lac-rbcS transcription fusion of pSS was subsequently ligated into pLl and generated a new construct, pLS23 (Fig. 4.2). Plasmid pLS23 enabled the rbcL and rbcS genes to be transcribed separately from identical promoters in a single host cell. In addition, expression levels of the two genes would not be affected by plasmid copy dosage since they were cloned in the same plasmid.

However, the rbcX gene was disrupted. The intact rbcLXS gene cluster was also cloned behind a lac promoter (pLS32) (Fig. 4.2).

When recombinant large and small subunits of RubisCO were synthesized in E. coli JM109(pLl) and E. coli JM109(pS5), respectively, each of the proteins was insoluble and precipitated after a crude lysate was centrifuged at 12000g. Solubility was recovered when these two subunits were coexpressed in the same cell using plasmids pLS23 or pLS32 (Table 4.1). RubisCO activity was barely detectable in the lysate of E. coli JM109(pLS23), whereas the lysate ofE. coli JM109(pLS32) showed an activity that was nearly 15-fold higher than E. coli JM109(pLS23). The activity was substantially increased when each plasmid was cotransformed with plasmid pKY206, a plasmid which Fig. 4.2. Restriction maps of subclones used for expression of Anabaena sp. strain CA rbcL, rbcS and rca genes in Escherichia coli. Details of the construction procedures are described in MATERIALS AND METHODS. Vectors (thin solid lines) used to clone the Anabaena DNA fragments (thick solid lines) are delineated beside each individual construct; each plasmid is labeled in boldface. The arrows indicate the transcriptional orientation of the Anabaena genes and the lac promoter (P lac) of each plasmid. Only one Dral site is shown in this figure and the restriction sites in small letters represent those that are inactivated after cloning. B=BamHl( D & d=DraI, E=£coRI, EV & ev=£coRV, H=///ndIII, hc=//mcll, Hp & hp=Hpal, N & n=A7id, P & p=PvwII and S & s^Sm al.

83 84

H HpHpEV HpH H E H D Hp Hp E

rbcL rbcX rbcS rca

I 1kb I K 1 8 / P/ac pK18::EV1.7-2 hc/ev Hp H H E H d/s

L y P K 1 9 / P/ac pLS32

hc/ev Hp H H E N

Z- L V p K 1 9 / P/ac pL1

B E H D Hp EV Hp E

/p R K 4 1 5 /_ _UL_L pRK415::E4.3

PB E Hd/s

z - h a l l L V P K 1 9 / P/ac pS5

hc/ev Hp H H E BE Hd/s

K 1 9 /

pLS23

Fig. 4.2 Table 4.1. Solubility and RubisCO activity of Andbanea sp. strain CA rbcL and rbcS gene products in Escherichia coli. The lysate of IPTG-induced cells of E. coli JM109 containing the indicated plasmid and its soluble extract obtained by a centrifugation at 12000g were analyzed by SDS polyacrylamide gel electrophoresis. Solubility was determined by the presence or absence of recombinant RubisCO subunit polypeptides in the soluble extracts. Considerable amounts of RubisCO subunit polypeptides were detected after SDS-PAGE of cell lysates prior to centrifugation. RubisCO activity was measured using the cell lysates. n.d.=not determined.

Plasmid | | PLl pS5 pLS23 pLS32 pLS23 pLS32 pKY206 pKY206

Solubility | -- + + + + RubisCO activity n.d. n.d. 0.2 2.8 7.8 13.8 nmoles/min/mg J 86 leads to constitutive and enhanced synthesis of the cpn60 (GroEL) and cpnlO (GroES) proteins of E. coli (116). The cpn proteins appeared to have a stronger impact on the pLS23 done, such that colransformation with pKY206 raised RubisCO activity from 0.2 to 7.8 nmoles C 02 fixed/min/mg protein (about a 40-fold increase), while the activity in cells containing plasmid pLS32 was increased from a specific activity of 2.8 to 13.8 nmoles C 02 fixed/min/mg protein (about a 5-fold increase) (Table 4.1). Since the lysate of E. coli JM109(pLS32 & pKY206) exhibited the highest activity, theAnabaena sp. strain CA RubisCO protein was purified from this recombinant clone.

In vitro characterization of RubisCO. Since RubisCO activase is thought to catalyze the removal of tightly bound phosphorylatcd metabolites that inhibit RubisCO activity (127), the role of such compounds in influencing RubisCO from Anabaena sp. strain CA RubisCO was examined. In contrast to the plant enzyme, Anabaena RubisCO exhibited a linear increase in activity during a 30-min time course (Fig. 4.3A). In other words, KABP and XuBP, which are formed by all sources of RubisCO, did not inhibit theAnabaena enzyme. Moreover, RuBP did not retard the conversion of the unactivated

Anabaena RubisCO into the ECM form. About half the carboxylase activity was recovered when the RuBP-bound enzyme (ER) was activated with 20 mM NaHC03 and

10 mM MgClj for 1 min; after 7 min of activation, the level of carboxylation was similar to the ECM form (Fig. 4.3B). In these experiments, 4 mM RuBP was included in the activation step to assure that the release of RuBP from the E form was not due to a change in the equilibrium after the volume increase. The difference in specific activity observed in Fig. 4.3A and Fig. 4.3B is presumably due to the use of different 87 (A) 100

60 oN u 40 s ■o 20

0 0 5 10 15 20 25 30 Reaction Time (min)

(B) 1.4 r 1.2 1

£ 0.8 c* O 0.6 u S 0.4 ER 75 ECM 0.2 0 _L _L 0 1 2 3 4 5 6 7 Activation Time (min)

Fig. 4.3. (A) Time course of Anabaena sp. strain CA RubisCO activity. Purified RubisCO enzyme was incubated with NaHN 03 and MgCl 2 for 5 min prior to initiation of the reaction with RuBP. At the indicated times, the reaction was quenched and acid stable ^ c product was determined. (B) Time-dependent activation of RuBP-treated Anabaena sp. strain CA RubisCO. The activity of RuBP-bound unactivated RubisCO (ER) was measured after preincubalion in the activation buffer for the indicated time periods as described in MATERIALS AND METHODS. The activity of the activated form of the (ECM) is also illustrated. 88 preparations stored for different times at 4°C. Preliminary results indicated that the activity of the Anabaena RubisCO was inhibited by incubation of the ECM form with

CA1P before initiation of carboxylation. Unfortunately, the level of inhibition was not consistent from time to time. However, the Anabaena enzyme appeared to be much less sensitive to this inhibitor than plant RubisCO. Due to the difficulty in obtaining enough of CA1P for thorough analysis, this part of the catalytic study was aborted.

Expression and purification of recombinant RubisCO activasc. Our previous study (98) reported partial purification of the recombinant Anabaena sp. strain CA activase. This purification has since been improved. It was found that RubisCO activasc separated from contaminating proteins when a 300 ml salt gradient (0-0.5 M KC1) was employed on the Q-sepharosc column previously employed (Fig. 4.4A). In addition,

HEPES was found to be more effective than Tris buffer at pH 7.0 and 0.2 mM ATP was included in the purification buffer to better stabilize the activase (146). The Anabaena sp. strain CA rca gene was expressed in E. coli at two temperatures, 37°C and 25°C.

The ATPase specific activity of the purified activase isolated from the 25°C culture was

4-fold greater than that exhibited by the activase purified from the 37°C culture (Fig.

4.4B). No effect of purified recombinant RubisCO activase was observed when this protein was added to CAlP-incubated Anabaena RubisCO.

Construction of a rca mutant. A genetic approach was also undertaken to investigate the physiological role of RubisCO activase in Anabaena. Although we have been unsuccessful thus far in transferring genetic information into Anabaena sp. strain

CA, there is a conjugation system available for A. variabilis and other Anabaena spp. Fig. 4.4. (A) SDS polyacrylamide gel electrophoresis analysis of RubisCO activase fractions after purification. Lanes 1 and 11 represent molecular weight markers (top to bottom): phosphorylase b, Mr, 97,400; bovine scrum albumin, 66,200; ovalbumin, 45,000; carbonic anhydrase, 31,000; soybean trypsin inhibitor, 21,500. Lanes 2 and 3 contain samples of the lysate and membrane-free extract of IPTG-induccd E. coli JM109(pK18::EV17-2) cells, respectively. Lane 4 contains a sample of the ammonium sulfate fraction, while the other lanes represent individual fractions after Q-sepharosc column chromotograghy. (B) Time course of ATPase activity by Anabaena sp. strain CA RubisCO activase. The purified activase was expressed in E. coli at both 25°C (■) and 37°C (□). ATP hydrolysis was quatitated by coupling ADP formation to NADH oxidization in an assay containing excess commercial pyruvate kinase and lactate dehydrogenase.

89 (A)

1 2 3 4 5 ( 7 3 I 14 1112 13 14 15 II 17 II II

Fig. 4.4

g Fig. 4.4 (continued) 4.4 Fig. (B)

Umoles NADH oxidized/mg 0.5 2.0 1.5 1.0 0 0 0.5 1 Time (min) Time 1.5 2 . 3 2.5 92 (42). Thus an A. variabilis rca mutant was constructed using this conjugation system.

TheA. variabilis rca gene, and the upstream rbcS gene were shown to be located within a 3.7-kb EcoRI fragment of RubisCO-containing plasmid 21C1 (75, 99). To inactivate rca, the gene was disrupted by deletion of an approximate 0.2-kb internal Clal fragment of the gene, followed by insertion of a Smr/Spf cartridge as a selective marker. The

£a?RI fragment containing the disrupted rca gene was then ligated into plasmid pNOT19

(159) along with another fragment containing oriT and sacB to create a mobilizablc suicide plasmid, pVA353 (Fig. 4.5A). With the assistance of helper plasmids pRL528 and pRL443 (42), pVA353 was mobilized into A. variabilis through conjugation.

The sucrose-inducible sacB gene from Bacillus subtilis encodes a levansucrase exoenzyme. Expression of this gene is lethal in Anabaena sp. strain PCC 7120 and various Gram-negative bacteria, probably because of fructosylation of essential acceptors or inadequate deposition of unsecreted levansucrase in cell membranes (20,53,54). Due to its inducible lethality, sacB has been exploited to select against single recombinants in these organisms (20, 53). In spite of the success obtained with Anabaena sp. strain

PCC 7120 (20), the addition of 5% sucrose into solid media did not help in the isolation of the A. variabilis rca mutant. On the contrary, sucrose hindered antibiotic selection by slowing the killing effect of the antibiotics on wild type cells and also initiated production of mucus, which in turn inhibited the formation of resistant colonies. The failure to select a double recombinant using sucrose may be related to the tolerance of

A. variabilis to sucrose (73). However, seven Smr/Spr clones were isolated by repeatedly streaking the resistant colonies on sucrose-free streptomycin/spectinomycin-containing Fig. 4.5. Inactivation of the Anabaena variabilis rca gene. (A) Construction of plasmid pVA353. The £coRI raz-containing DNA fragment of Anabaena variabilis (thick line) was First cloned in pK19 to generate pVA19K. The rca gene was modified by replacing the internal Clal fragment with a Smr/Spr cartridge to form pVA64K. The cloning vector was then switched to pNOT19 (pVA64N) for convenience of subsequent insertion of a Notl fragment containing oriT and the sacRsacB opcron. B-BamUl, C=Clalt E=£coRI, N=Aro/I, Apr=ampicillin resistance, Kmf-kanamycin resistance, Smr=streptomycin resistance, Spr=spectinomycin resistance, Tmr=Trimethoprim resistance. Plasmid pVA353 was then mated inio Anabaena variabilis and Smr/Sprclones were examined by Southern blot hybridization analysis using the Anabaena sp. strain CA rca gene (B) and the Smr/SpT cartridge (C) as probes. Lane 1, plasmid pVA353; lane 2, genomic DNA isolated from wild type A. variabilis ; lanes 3 to 9, DNA isolated from 7 A. variabilis transformants. In all cases, DNA was digested with EcoRI prior to hybridization.

93 pVA19K

pVA64K

pVA64N

pVA353 95

Fig. 4.5 (continued)

C 96 plates. Southern blot hybridization analysis showed that the size of the rca-bearing

£coRI fragment from all seven clones was larger than that from the wild type strain, but similar to that of the donor plasmid (Fig. 4.SB). The increase in size was the result of the insertion of the Smr/Spr cartridge, since DNA isolated from these 7 clones all hybridized to the cartridge probe (Fig. 4.5C). One of the difficulties in mutagenizing filamentous cyanobacteria is the potential existence of both wild type and mutant cells within a filament. Thus, it has been suggested to gently break the Anabaena filaments into single cells or short fragments by mild sonication before conjugation to facilitate homogenous mutation (186). During the course of constructing the rca mutant, we found that sonication decreased the cffcciency of conjugal transformation. In fact, the most effective means of isolating a homogeneous mutation was by repeated streaking on selective plates.

Characterization of the rca mutant. Southern blot hybridization repeatedly demonstrated that the rca mutant, strain 353, was stable when grown in the absence of streptomycin and spcctinomycin selection. Therefore, the mutant strain was characterized under such conditions in order to be directly comparable to the wild type strain. The wild type and mutant strains grew at similar rates when 1 % C02 in air was supplied to the cultures; the doubling time under low, medium and high light intensity was 15.4, 10.5, and 7.1 h for the wild type strain and 14.9, 9.6, and 6.7 h for strain

353, respectively. Interestingly, when shifted from low light to high light at 1 % C02 in air, the wild type strain immediately increased its growth rate while the mutant strain maintained a similar growth rate for at least 12 h (Fig. 4.6A). In addition, the wild type 97 (A).

A 660

O WT/LL-*HL • WT/LL A 353/LL->HL 4 353/ LL 0.1 0 2 4 6 8 10 12 Time (hr) (B)

A 660

O WT A 353

0.1 0 10 20 30 40 50 60 70 80 Time (hr)

Fig. 4.6. Growth response of Anabaena variabilis wild type and rca mutant strain 353. (A) At mid-logarithmic phase, wild type and mutant cultures previously grown at a light intensity of 8 pmoles quanta m-2 sec-1 and bubbled with 1 % C02 in air were divided into two equal portions. One half of each culture was shifted to 73 nmoles quanta m*2 sec*1 (LL-> HL) and the other half remained under the low light intensity (LL). (B) The wild type and mutant cells were bubbled with air under a constant illumination of 73 quanta m*2 sec'1. 98 strain grew about 1.4 times faster than the mutant at air-levels of C02 under high light; the doubling time was 44.5 h for the wild type and 62.2 h for the mutant (Fig. 4.6B).

The air/high light cultures of both strains entered stationary phase around A**, of 0.7 while the 1% C02 cultures exhibited exponential growth up to Amo of 2.5 under high light and to an Amo I under low light. An unusually long growth lag was observed when strain 353 was inoculated into the strcptomycin/spectinomycin-containing media and cultured under limiting C02/high light conditions; sometimes the inoculum cells lysed and no growth was initiated. By contrast, the mutant grew relatively well in the presence of the antibiotics under 1% C02/high light or air/low light conditions.

Under low light, the 1% C02 cultures of the wild type and mutant strains displayed similar levels of whole-cell RubisCO activity when several separate assays with different cultures were averaged (Fig. 4.7A); relatively small differences between these two strains were sometimes observed but were considerably less than those obtained under high light or tow C02 conditions (Fig, 4.7B, 4.7C and 4.7D). The activity exhibited by the air/low light culture of the wild type strain was comparable to the activity of the 1% C02/low light culture but about 30% less carboxylation over a 10-min assay was detected for the air culture of the mutant strain (Fig. 4.7B). When bubbled with 1 % C02 under high light, the level of RubisCO activity in the wild type strain increased substantially while the activity of the mutant paralleled the activity of mutant cells incubated under low light (Fig. 4.7C). Although there was no obvious difference in the pattern of activity exhibited by air-grown cultures of each strain at high and low light intensities (Fig. 4.7B and 4.7D), differences in the level of carboxylation exhibited Fig. 4.7. WhoIe*ccll RubisCO activity of Anabaena variabilis wild type and rca mutant strain 3S3. Both strains were cultured under 4 different conditions: 1 % C 02 in air/8 /nmoles quanta m*2 sec1 (A); air/8 /xmolcs quanta m'2 sec'1 (B); 1 % C02 in air/73 /imoles quanta m'2 sec'1 (C); and air/73 /nmoles quanta m’2 sec'1 (D). Carboxylation was initiated adding toluenc-permeabilizcd washed cells to each reaction mixture. At the indicated time points, a sample from each cell suspension was acidified to terminate the reaction and the amount of I4C product was determined by liquid scintillation spectrometry. Each experiment was repeated at least once. The curves obtained from individual assays were analyzed using computer to generate equations of curvature. The equations for each given growth condition were averaged and the value deduced from the average equation for each time point was drawn.

99 I

(imoles CO 2 fixed/mg (imolcs CO 2 flxed/mg 0.5 2.5 . •• 1.5 ■■ 2.5 0.5 . ■■1.5 2 3 2 3 0 1 1 ■■ ■ ■ ■ • (A ) 8 8 ) (A (C ) ) (C a i f 3noe -1 s ^ m nmoles 73 /AI 2 0 C % 1 2 0 C % 1 nmoles nmoles 8 A 6 8 6 4 O Time (min) Time m*2 s-1 m*2 s-1 /AI t t A 8 O i f ^ WT o 10 353 Fig. 4.7 Fig. (B ) 8 nmoles m"2 m"2 s"1 nmoles 8 ) (B (D ) ) (D 8 Air Air 73 nmoles nmoles 2 A 6 8 6 4 o Time (min) Time m*2 A A o o s"1 A o 0 12 10 A o ■ 3 -■ ■■ ■■1.5 ■■ ■■1.5 ■■ ■■2.5 ■■0.5 ■■ ■■2.5 ■■0.5 2 0 2 1 1

101 by the mutant and wild type were accentuated under air/high light conditions (Fig. 4.7D).

The lower carboxylase activity exhibited by strain 353 appeared not to result from lower

RubisCO content since Western blot analysis indicated that RubisCO levels in the wild type and mutant were similar for each condition examined. The upward curvature of the activity response curve in figure 4.7 is obviously due to a time-dependent activation of the enzyme in whole cells during the assay. To monitor the response of both the mutant and wild type strains at the different light regimes at air-levels of C 02 , the specific activity for each time interval was plotted as a function of the reaction time (Fig. 4.8).

In these experiments, it was apparent that RubisCO in the mutant cells was not able to reach the maximum level of activity obtained by the enzyme in the wild type background; in addition, maximum activity was achieved much more rapidly at high light intensities for both the wild type and mutant. The initial rates of the specific activity increase were considerably diminished in the mutant, 25 mU/mg per min under low light and 30 mU/mg per min under high light, while the rates were 38 and 55 mU/mg per min for the wild type at low light and high light intensities, respectively. It is also apparent that this effect is enhanced when cells are incubated under high light intensity at low levels of exogenous C02.

DISCUSSION

The large and small subunits of Anabaena sp. strain CA RubisCO sedimented with cellular debris after crude lysates of separately synthesized recombinant subunits 102 (A) Air 73 jimoles quanta m"^ s"l 0.25 -r

0.2 - -

0 .1 5 --

0.1 - -

g 0 .0 5 --

0 2 4 6 8 10 12 Time (min)

(B) Air 8 jim oles quanta m"2 S"1

0.25-r- cm | 0.2 -

I 0.15-7

~~10 15 20 25 30 35 Time (min)~~

Fig. 4.8. Time-dependent increase of RubisCO specific activity of Anabaena variabilis wild type and rca mutant strain 353. Whole-cell RubisCO activity of both strains grown under the indicated conditions was measured after incubating permeablized cells with the reaction solution for the indicated time intervals; specific activity was plotted as a function of reaction lime. 103 were prepared in E. coli. These results indicated that subunit monomers formed improperly assembled, insoluble aggregates in the absence of the companion subunit polypeptide. A similar phenomenon was observed when attempts were made to express therbcL gene of Zea mays in E. coli (52). In addition, when large subunits of Anabaena sp. strain PCC 7120 RubisCO were synthesized in great excess relative to small subunits in E. coli, the extra large subunits also became insoluble (70). By contrast, the large and small subunits of Synecbococcus sp. strain PCC 6301 RubisCO retained most, if not all, of their solubility when synthesized independently in E. coli cells. In cell extracts, a majority of the recombinant Synecbococcus large subunits were assembled into Lg octomcrs while the small subunits were found predominantly in the form of free monomers. Active holoenzymes could be reconstituted spontaneously by mixing crude preparations of each subunit (4, 96, 122). The differences in solubility suggest that interactions between large and small subunits favors an equilibrium where assembly of the holoenzyme is the favored pathway, thus stabilizing the LgSg structure and its assembly intermediates. This scenario appears to be much more pronounced for the

~~Anabaena RubisCO than theSynecbococcus enzyme. It appears that several conserved amino acid residues of the Anabaena sp. strain PCC 7120 and Synecbococcus sp. strain~~

PCC 6301 small subunits may be important for the formation of a soluble and functional holoenzyme (47, 94). It is also possible that the Anabaena and Synecbococcus RubisCO holoenzymes are assembled via diverse pathways, although there is no direct evidence to support this idea. Recent studies with the Synecbococcus sp. strain PCC 6301

RubisCO suggest that the large subunits first form Lj dimers; the Lj dimers are subsequently tctramcrizcd into L* octomers; finally, small subunits spontaneously associate with the L, core to generate the L,Sg structure (4, 68, 96, 122). It was also suggested that the Anabaena small subunits might be involved in assembly at earlier steps

~~(70). The other possibility is that other components are required for RubisCO assembly.~~

~~Indeed, newly synthesized plant large subunits were found in association with the~~

~~RubisCO binding protein and remained soluble in chloroplast extracts before assembly~~

(150). Chloroplast RubisCO binding protein or its bacterial homologue, the cpn60 protein (GroEL), together with the cpnlO protein (GroES), were shown to promote the fomation of L, dimers from unfolded monomers (67, 68).

~~There is an open reading frame, rbcX, situated between rbcL and rbcS in~~

Anabaena sp. strain CA (99) and Anabaena sp. strain PCC 7120 (93). Elimination of expression of the putative rbcX gene, either by independent transcription of rbcL and rbcS from two lac promoters on one plasmid (pLS23 of this study) or by deletion of rbcX from the operon (93), decreased the production of active RubisCO. Moreover, RubisCO activity increased 41-fold when pLS23 was cotransformed with the GroELS-encoding plasmid pKY206. Lack of expression of rbcX appeared to demand the presence of a large excess of chaperonin proteins for the formation of the LgSg structure. These results imply that the product of rbcX might play some role in the assembly of Anabaena

RubisCO. Unlike the RubisCO-binding protein, the deduced amino acid sequence of rbcX showed little similarity to known chaperonin proteins. Obviously, much more work is required to determine if the gene product of rbcX is actually involved in C02 fixation, but the preliminaiy results thus far generated are certainly suggestive. The key position 105 of rbcX in the rbcLXS operon presumably is not fortuitous.

~~Biochemical studies with recombinant Anabaena sp. strain CA RubisCO indicated that the regulatory properties of the Anabaena enzyme were quite different from plant~~

~~RubisCO. The E form of Anabaena RubisCO was easily dissociated from RuBP when~~

~~HCO,‘ and MgI+ were provided, thus allowing activation to proceed rapidly. Failover is thus not a phenomenon that appears to influence the activity of the Anabaena enzyme.~~

Some degree of inhibition was demonstrated by preincubation with CA1P; however, the inhibition was not so significant as that imposed on the plant enzyme. Despite the fact that we do not have a method to assay activase-mediated RubisCO activation in vitro at this time, it is clear that the specificity ofAnabaena RubisCO activase appears to diverge from plant activasc. Interestingly, the immunity to RuBP inhibition is also observed with the L*Sg RubisCO enzymes from other cyanobacteria (5, 95) and the type II enzymes from RJiodospirillum rubrum and Rhodobacter sphaeroides (179). In addition, it was found that nonchlorophytic eukaryotic algal RubisCO enzymes exhibit a varied ability to be inhibited by RuBP (119, 138). At this time, it is not clear whether nonchlorophytic eukaryotic algae contain RubisCO activase. On the other hand, the distribution of

RubisCO activase in cyanobacteria is controversial. Our previous studies indicated that unicellular and nonheterocystous filamentous cyanobacteria lack rca-Iike genetic elements when probed with plant and Anabaena rca sequences at high and low stringency (98).

~~Another group demonstrated, using immunogold lebeling with antibodies to plant~~

RubisCO activase, that the enzyme was uniformly present in the cytoplasm of the unicellular cyanobacterium Synecbococcus sp. strain PCC 7942 (48). In this latter study, 106 no preimmune antiscra was included as a control. However, variations in the regulatory * properties of RubisCO certainly imply diverse mechanisms to modulate C02 fixation among different species.

The rca mutant of A. variabilis, strain 353, exhibited a slower growth rate and a lower level of whole-cell RubisCO activity than the wild type when cultured at air- levels of C02 (Fig. 4.6B, 4.7B, 4.7D and 4.8). A high C02 requirement for growth was also observed with activasc-deficient plants (110, 170) and it was found that RubisCO activase circumvented the demand for increased levels of C02 by enhancing the affinity of RubisCO for activator C02 in chloroplasts (128). Cyanobacteria! RubisCO proteins usually gather in the carboxysomes, polyhedral subcellular structures where C02 is concentrated locally (131, 132). Displacing the RubisCO-encoding genes of the cyanobacterium Synechocystis sp, strain PCC 6803 with a bacterial Type II (L,) RubisCO gene resulted in no carboxysomes and a high C02-requiring phenotype in the mutant (2,

123). Due to the enhanced effect on RubisCO activity in mutant strain 353, it will be interesting to determine if the mutation in the Anabaena rca gene leads to an alteration in structure or number of carboxysomes in this organism.

~~RubisCO activase appeared to be important for the elevated RubisCO activity exhibited by wild type A. variabilis under 1% C02/high light conditions (Fig. 4.7C).~~

The correlation between high light intensity and the requirement for RubisCO activase was also indicated by the fact that the 1% C02 culture of strain 353 was not able to accelerate growth immediately after being shifted from low light to high light (Fig.

4.6A). The initial rate by which RubisCO is activated to its maximum specific activity 107 increased under high light conditions; however, the increase observed for the mutant was less than that for the wild type (Fig. 4.8). It is believed that the active synthesis of

RuBP during illumination stimulates RubisCO deactivation in activase-deficient plants by preferential binding to the E form, thus switching the activation equilibrium toward decarbamylation (110,128,154). Although the carbamylation level of RubisCO was not measured in this study, RuBP is not likely the cause for the relative inability of the

~~Anabaena rca mutant to parallel the RubisCO activity of the wild type under 1%~~

~~COz/high light conditions since this metabolite docs not tightly bind unactivatcd~~

~~Anabaena RubisCO, at least at the concentrations employed in this study. Accordingly,~~

~~XuBP and KABP, the isomers of RuBP, arc probably not obstacles for efficient C02 fixation in cyanobacteria. Thus, if metabolic inhibitors influence the light regulation of~~

~~RubisCO activity in Anabaena species as in plants, distinct compounds may be involved.~~

Certainly, at this stage, we can not exclude the possibility that RubisCO activase regulates RubisCO activity via some other sort of mechanism. It is apparent that the function of RubisCO activase in cyanobacteria, especially in Anabaena species, is somewhat unique from the theme established in plants. CHAPTER V

~~Cyanobactcrial Genes Encoding Phosphoribulokinase and Ribulose Bisphosphate~~

~~Carboxylase/Oxygenase Arc Not Closely Linked~~

~~INTRODUCTION~~

~~The Calvin reductive pentose phosphate pathway is the primary route to convert~~

C02 into organic carbon for a wide range of autotrophic organisms, including cyanobacteria. The two successive reactions catalyzed by phosphoribulokinase (PRK) and ribulose bisphosphate carboxylase/oxygenase (RubisCO) are essential and unique to this pathway. PRK catalyzes the ATP-dcpendent phosphorylation of ribulose 5-phosphate to yield ribulose 1,5-bisphosphate (RuBP), which is in turn carboxylated via RubisCO catalysis. It has been shown that the structural genes encoding PRK, RubisCO, as well as other Calvin cycle enzymes, arc clustered in a variety of autotrophic bacteria (61, 77,

~~112). These genes are coordinately regulated by a LysR-typc of transcriptional activator,~~

~~CbbR (44, 62, 183, 190, 203).~~

In contrast to the wealth of research on RubisCO, there is a paucity of knowledge regarding PRK. However, it is clear that plant PRK enzymes are very different from the enzymes of many anoxygenic photosynthetic and aerobic chemosynthetic microorganisms, particularly relative to aspects of enzyme structure and activity regulation. Plant PRK

108 109 enzymes are homodimers and are subject to regulation through reversible reduction of a disulfide bond by a fcrrcdoxin/thioredoxin system, where reduction/oxidization is tightly linked to the photocyclc. By contrast, bacterial PRK exhibits a multimcric structure of identical subunits and many enzymes require NADH for maximum activity

~~(40). The PRK-cncoding genes from several sources have been sequenced (57, 58, 88,~~

~~102, 112, 113, 147). The deduced amino acid sequences of the plant and green algal enzymes share little homology with the bacterial proteins.~~

~~Early studies with cell extracts of Anacystis nidulans (now classified as~~

~~Synecbococcus sp. strain PCC 6301) (37) showed that cyanobacterial PRK was activated by light in vivo and in vitro by disulfide reduction after incubation with dithiolhreitol~~

~~(DTT). Studies with the enzymes prepared from Chlorogioeopsis Jritschii (109) and~~

Anabaena cylindrica (161) further confirmed the significant effect of disulfide reduction on the activation of cyanobacterial PRK. Although DTT stimulated the activity of both enzymes, a much stronger effect was demonstrated by DTT-reduced thioredoxin on the

~~A. cylindrica PRK. On the other hand, NADH did not regulate PRK activity in cell extracts of Anabaena sp. strain CA and Agmenellum quadmplicatum (now named~~

~~Synecbococcus sp. strain PCC 7002) (177) and C. fritschii (109).~~

The structure of cyanobacterial PRK is controversial. The enzyme from C. fritschii was reported to contain six identical monomers and the molecular weight of the monomer was estimated to be 40,000 (109). The A, cylindrica enzyme was shown to have a native molecular weight of 72,000 and SDS polyacrylamide gel electrophoresis yielded nearly equivalent amounts of two polypeptides, of Mf 43,000 and 26,000 each, no although the small polypeptide was speculated to be a tightly bound contaminant (161).

Based on their catalytic and regulatory properties, cyanobacterial PRK enzymes appear to more closely resemble plant enzymes than the enzymes from other prokaryotic sources. As part of our interest in discerning the organization and regulation of the

~~Calvin cycle genes in cyanobacteria, we decided that the spinach prk gene would be a suitable probe to help identify the prk gene of Anabaena sp. strain CA.~~

~~MATERIALS AND METHODS~~

~~Bacterial strains and plasmids, Escherichia coli HB101 (17) was used as the host for recombinant derivatives of cosmids pVK102 (86) and pLAFRl (SO) while E. coli~~

JM109 served as the host for plasmids pUC18, pUC19 (207) and pRK404 (35). The cosmids were used for the construction of a genomic library. The plasmids pRK404 and pUC18/19 were the vectors employed for subcloning and sequencing. A 1.4 kb spinach cDNA fragment, cloned in the EcoRI site of pBR325 (113), was used as a heterologous probe in these experiments.

~~Media and growth conditions. Anabaena sp. strain CA (ATCC 33047) was grown in ASP-2 medium (189) containing 10 mM NaNOj at 39°C and bubbled with 1 %~~

COj/99% air. E. coli cells were cultured at 37°C in Luria Broth (33) containing the following supplements when necessaiy: ampicillin, 100 /ig/ml; kanamycin, 50 /xg/ml; tetracycline, 12.5 fig/ml; agar, 1.5% (w/v);5-bromo-4-chloro-3-indolyI-B-D-galactoside,

40 /ig/ml; isopropyl-B-D-galactoside, 0,1 mM, Ill DNA manipulations. Anabaena DNA was extracted using the sarkosyl lysate procedure (46) and further purified through a cesium chloridc-cthidium bromide gradient.

~~The alkaline lysis method (13) was used to isolate plasmid DNA from E. coli. Enzyme digestions were performed using conditions recommended by the suppliers. Recombinant~~

DNA was transformed into CaCl2*trcatcd E. coli cells following the protocol described by Maniatis et al (106). Nucleotide sequences were determined by the didcoxy chain termination method (156) using a Sequcnase kit (U.S. Biochcmicals).

Gene library construction. A partial Anabaena sp. strain CA gene library was constructed using cosmid pLAFRl. Genomic DNA of Anabaena sp. strain CA was partially digested with EcoRI for different time intervals to yield fragments of about 20 kb; the 5' phosphate groups of the DNA fragments were then removed with calf intestinal phosphatase. The dephosphorylated DNA fragments were inserted to the £coRI site of pLAFRl and the recombinant cosmid molecules were transfected into E. coli

HB101 with the assistance of lambda particles. Approximately 900 colonies were obtained. A second Anabaena sp. strain CA gene library was constructed using cosmid pVK102 as the cloning vector. Procedures for the construction of the pVK102 library were described in CHAPTER II (98).

Hybridizations. For colony hybridizations, cells of individual clones were grown and later lysed on nitrocellulose membranes (106). DNA released from the lysed cells was fixed onto the nitrocellulose membranes by baking the blots at 80"C under vacuum for 2 h. The blots were hybridized with a 1.4-kb spinachprk cDNA probe at 52°C and rinsed several times after hybridization with 5X SSC (IX SSC: 0.15 M NaCl, 0.015 M 112 sodium citrate, pH 7.0) plus 0.1% sodium dodecyl sulfate (SDS) at room temperature followed by two 1 h washes with 2X SSC-0.1% SDS at 40°C. For Southern blot hybridization, DNA fragments separated by gel electrophoresis were transferred to

~~GeneScrccn Plus membranes by a capillary method, described in a protocol distributed by the membrane manufacturer. Hybridization to the spinach probe was performed at~~

~~47°C and the blots were washed four times with 5X SSC-0.1% SDS at 37°C for 30 min for each wash.~~

~~RESULTS AND DISCUSSION~~

~~An attempt to identify the prk gene of Anabaena sp. strain CA was made through~~

~~Southern blot hybridization analysis of genomic DNA using a spinach prk cDNA probe.~~

~~The probe hybridized with a 6-kb //rndlll fragment or a 9-kb EcoRI fragment (Fig.~~

5.1A), Clones containing the hybridizing region were isolated from two cosmid libraries of Anabaena sp. strain CA DNA and the 6-kb //mdlll fragment was subcloned into pRK404. Further analysis of the //mdlll fragment showed that hybridization occurred within a 0.8-kb EcoRV-Hpal fragment (Fig. 5. IB), The 0.8-kb fragment and its flanking regions were sequenced; however, no PRK-encoded open reading frame was found (Fig. S.l and 5.2). A search for similarities between the 0.8-kb Anabaena DNA and the spinach prk cDNA, using the BESTFIT program, revealed that only a 17- nucleotide stretch on one strand and a 25-nucleotide stretch on the other strand of the

Anabaena fragment possessed substantial levels of identity to the spinach DNA sequence Fig. 5.1. Southern blot hybridization using a spinach prk cDNA probe. (A) Lanes 1 and 2 depict the EcoRI and tfmdlll digested genomic DNA of Anabaena sp. strain CA, respectively. (B) A clone containing the 6*kb 7/mdIII fragment of panel A was digested with EcoRV and Hpal (lane 1), with Hpal (lane 2) and with EcoRV (lane 3). (C) Restriction map of the 6-kb 7/mdIII fragment that hybridizes to the spinach prk probe. The box indicates the region where hybridization occurred. The arrow shows the region which has been sequenced and the direction in which the sequence is presented (Fig. 5.2). E=EcoRI, EV=EcoRV, H=Hindlll, Up=Hpal.

~~113 114~~

~~c. EV/Hp Hp/EV~~

~~H Hp EV Hp EV Ev \/ e EV H~~

~~lkb~~

Fig. 5.1 Fig. 5.2. Nucleotide and deduced amino acid sequences of the region of Anabaena strain CA genomic DNA hybridizing to a spinach prk probe. The deduced amino acid sequence is listed beneath the nucleotide coding sequence in the single letter format; the second methionine residue in the open reading frame is highlighted in boldface. Restriction sites for EcoKV and Hpal are underlined. Letters N and X represent ambiguous nucleotide and amino acid residues, respectively.

~~115 116~~

CAAATACCTG AAATATACAT TTCCCAAGCA TAGAGATGTG ATTGCAGAAG TTTATCAACT TCCCTTAATG GATAACATAA TTACAGAAAG AAGTGTCTTG 100 TCACATTGTA TTAATAAAAC TGTAGCTATA GGTAATTTGC CAGTACCTGA TCCTAAATCA ATTTTTGAGT TTGCAAAGTA TGACTCATCT GCAAGTCAAT 200 CAGCTTTTGA ATTTGAGTCA TTAGCACTAG AAGTCAGAAG AAAACTAGAG ATACATTAAT GATTGATTTT CTGATGCTGA TAAATATAGA AAAGTTGAAA 300 AAACACGCCT TGTGATTTTT GCTCACTTAA ATAGGCGGAA AAATTGTCAT TAGGGTGATA TCACACTTTA CACCAGTTGC TTGATTTTGA TTAAAGTATG 400 GCAGCAAAGT ATACATTTTA TATTAAGAAA ATTTAATGTA ATAGCTGGTA TAACTTCTCG TTTTATTTAG GAATAAATTT AATTTAATTG TGTAAAATTA 500 CTTAATATAA TTTCATTACT ATACTTACTA TTAAGCAATT CTTATAGCTT AAACGGACTA GCTTATGTTA CAATAAATAC AAATAGACCT TTTTCAAATT 600 TTATAGAGGT AAAAAATTTG TTGTTTAATC CTGAGCTATG CCGTAACGAA AGCGAAGTCG AAAGCAAGCT CATTGTACAG TATTTACTAC CGCAGCTAGG 700 TTACACACCA GACACCTGGC ATCAGGAAGT GGCTGTTGGT AGCATCCGGT TAGATTTTTT AGCCTTTGCT GTACAAGTCG TTCCCTTAGT CTTAGATGCC 800

~~AACTCGCCTT TGAGTGTGGT A ATG GAA GCA AAG CAC CCC AAA 842 MEA K HPK~~

~~CAAAATTTAAATCAT CAT GTCCCTAGA CTC AGGCATTAT 881 Q NLNH HV PRL R HY~~

~~TTA ACA AGTTTG AAT GTAGGCTAT GGTTTA TTAACTAAT 920 LTS L N V GYGLLTN~~

~~GGTAGGGAAATCAGG ATTTAT GAAAAAGTT AAT AGTGAT 959 G REI R IYE KV NS D~~

~~GTTAAGCTAGTTTTT CAGTGTGCTGGC GAG GAAGTTGAG 998 V KLVF Q CAG E EVE~~

~~GCACATTTAGAAAAT ATT AAAGATTTA ATT GGCAGAGAC 1037 AHLEN I KDL I GRD~~

~~AGTCTTAAATAT AGG AAAATT ACCAACAAA CCA GAAACA 1076 S L KYRKITN K PET~~

~~TCAGTAAATTCAGAA AATATCACCGTT AAA GGAAAGGGG 1115 SV NSE N ITVK G KG~~

~~CAGCATTCAATG AAA ATCATC GCAGTCTAC CACAATAAA 1154 Q HSH K I IAVYHNK~~

~~GGCGGAGTTGGT AAG ACAACA ACTGTA GTT AAC CTAGCA 1193 GG VGK T TTV V NLA~~

~~GCCGCTATC AGAANA AAAGGT AAA CGAGTA TTAGTTATT 1232 A AIRXKG KRV L VI~~

~~Fig. 5.2 117 Fig. 5.2 (continued)~~

~~GATTTAGATAGCCAA GCAAATACTACATTT GCT ACT GGT 1271 DLDS Q A NTTF ATG~~

~~TTAGTTAAATTT GAT GATGAAGAATTTGAT GAT ATC AAA 1310 L VKFD DEEFD D I K~~

~~GACTCTAAT ATTTTT CATGTCTTATCATCA GAA GAC TTT 1349 DSNI F HVLSS E D F~~

~~TAT CCTATTTCGGAA GTTGCTAGAAAGTCT AAT TTT AGC 1388 YPI SE VARKS NFS~~

~~AAT CCAGAAATTTAT GTGATTCCC TCGCAC ATA GAT TTG 1427 NPEIY V IPSH I D L~~

~~ATA AGA CAA GAAAAT GAACTGAAT NCA CTT GAA TAT AGC 1466 IR Q EN ELNX L E Y S~~

~~AAACTAATATTAATT GAA AAACTAGAGGAA GTA AAA GAA 1505 KLIL I EKLE E V K E~~

~~GAATATGATGTTGTT TTGATAGATACACCG CCG TCT TT 1543 EY DVV LID TP P S 118 (Fig. 5.3A).~~

TheAnabaena nucleotide sequence obtained thus far contains an incomplete open reading frame, of which the C-terminal sequence remains unknown (Fig. 5.2). The deduced amino acid sequence immediately following the second methionine of the putative open reading frame shows significant homology to the N-temrinal regions of the iron protein subunits of protochlorophyllidc reductase, chlorin reductase and nitrogenasc

~~(Fig. 5.3B). Protochlorophyllidc reductase converts protochlorophyllide a into chlorin~~

(chlorophyllidc a) in the light-independent biosynthesis of chlorophyll and bacteriochlorophyll, and chlorin is further transformed into bacteriochlorin by chlorin reductase in the bacteriochlorophyll biosynthetic pathway, while the iron protein component of the nitrogcnase complex binds ATP and catalyzes the reduction of the molybdo-iron protein component of the complex, which in turn fixes N2 into NH} (45).

~~Each of the three enzyme complexes is composed of three different subunits and one of the subunits is an iron protein. These iron proteins all contain a site characteristic of~~

ATP binding (76, 192) at their N termini and two aspartate residues postulated to be involved in ATP hydrolysis (55) are conserved not far downstream from the ATP-binding site (19). Perhaps the Anabaena open reading frame also encodes an ATP-requiring protein, although the Anabaena sequence is missing the first residue of the consensus

~~ATP-binding sequence motif (G/A)XXXXGK(S/T) and one of the important aspartate residues is displaced by an asparagine.~~

PRK is also an ATP-binding protein and possesses a consensus ATP-binding sequence, which starts at a position 12 amino acids downstream from the leader peptide Fig. 5.3. (A) Nucleotide sequences with a high percentage of identity to spinach prk. Two short fragments (top line of each pair) within the 0.8-kb EcoRV-Hpal fragment shown in Fig. 5. IB exhibit 88% identity to the spinach prk cDNA sequence (bottom line of each pair). Anabaena nucleotide sequences are numbered consistent with those depicted in Fig. 5.2. (B) N-tcrminal amino acid sequence comparison. The deduced sequence which begins at the second methionine of the Anabaena sp. strain CA open reading frame (OrM/wCA) (Fig. 5.2) is compared with the N terminus of the iron protein subunit of protochlorophyllidc reductase (ChlL and BchL), chlorin reductase (BchX), or the nitrogenasc complex (NifH) from the following species: Synecltocystis sp. PCC 6803 (Cyny3), Chlamydomonas reinhardtii (Chlre), Pinus contorta (Pinco), Rhodobacter capsulatus (Rhoca), Rhodobacter sphaeroides (Rhosh), Anabaena sp. PCC 7120 (/Iflasp) and Azospirillum brasilense (Azobr). The asterisks indicate identical residues. The elements suggested to be involved in binding and hydrolysis of ATP are highlighted. The number of amino acids (aa) between dashes indicates the length of nonhomologous deduced amino acid sequence; gaps, marked by dots, are introduced in order to illustrate the most favorable alignment.

~~119 936 ATTTATGAAAAAGTTAA 952 II II IIII ! I I I I 1 I ATGTATGAACAAGTTAA~~

~~X084 TTTACTGATGTTTCTGGTTTGTTGG 1060 I I! I I I I It I I I I I I I I I ! I I I TATAATCATGTTTCTGGTTTGTTGG~~

* * * * * * Orf--AnaCA MK ...... IIAVYH WNLAAAIRK KGKRVLVIDL JSQANTT. FA ChlL-syny3 MT--- 2 aa---LAVY. SCNISTALAK RGKKVLQIGC JPKHfiST.FT ChlL-Chlre MK ...... LAVY. SCNISIALRK RGKKVLQIGC 3ST.FT ChlL-Pinco MK ...... IAVY. SCNISVALAR RGQKVLQIGC 3ST.FT BchL-Rhoca MS— 36 aa FSVY. SSNLSAAFSL LGKRVLQIGC |ST. FT BchX-Rhoca MT--34 aa--IIAIY. LANLSHMMAE MGKRVLLIGC TSLL BchX-Rhosh MT--34 aa— VIAIY. LANLSYMMAQ IGKRVLLIGC fSLL NifH-Anasp MT--- 6 aa---IAFY. SQNTLAAMAE MGQRIMIVGC STRLM NifH-Azohr MS--- 3 aa— -IAFY. SQNTLAALVE LDQKILIVGC STRLI NifH-Rhoca MG--- 4 aa---IAFY. SQNTLAALVE MGQKILIVGC STRLI 121 cleavage site of the spinach enzyme (113,125) (Fig. 5.3A). The similarity between the

ATP-binding sites possibly contributes to the hybridization between the 0.8-kb Anabaena fragment and the spinach probe; this is in addition to the two homologous DNA stretches found using the BESTFIT computer program. The prk gene has recently been cloned from the cyanobacterium Synechocystis sp. strain PCC 6803. The Synechocystis PRK sequence exhibits notable identity (61%) with the spinach enzyme (175). The difficulty in identifying the Anabaena prk gene using a heterologous plant probe perhaps reflects the fact that theAnabaena enzyme is less similar to the plant PRK than theSynechocystis enzyme.

It is apparent that prk is not linked with the genes encoding the large and small subunits of RubisCO ( rbcL and rbcS) in cyanobacteria as found in nonsulfur purple photosynthetic bacteria and chemolithoautotrophic bacteria (44, 61, 77, 112, 183).

Cosmid clones containing Anabaena sp. strain CA rbcLrbcS and the flanking regions (98) did not hybridize with the spinach prk probe in Southern blot hybridization analysis. In addition, the approximate 3-kb region immediately downstream from Anabaena sp. strain

CA rbcLrbcS has been sequenced (98, 99) and no Calvin cycle structural genes are encoded within this 3-kb region. Moreover, the regions 5* and 3' to the rbcLrbcS genes of Synechococcus sp. strain PCC 7942 also do not contain structural genes of the Calvin cycle (49, 101, 158).

~~The regions flanking the rbcLrbcS genes in cyanobacteria are important for C02 metabolism despite the fact that these sequences are completely different in heterocystous~~

Anabaena strains and unicellular Synechococcus strains. In Anabaena sp. strain CA, a 122 RubisCO activasc gene ( rca), preceded by two unknown open reading frames, was located downstream from rbcS (98). Evidence was presented to indicate that the gene immediately 3' to Synechococcus sp. strain PCC 7942 rbcS encods subunit II of bacterial phosphoribosyl aminoimidazole carboxylase (purK). Transcription of the Synechococcus purK gene was induced by low C02, and inactivation of the gene by deletion or insertion resulted in strains with a high C02 requirement for growth (158). The high C02- rcquiring phenotype was also observed when mutations were isolated in four open reading frames situated between 4.8 and 1.2 kb upstream of the Synechococcus rbcL gene; these are the ccmK, ccmL, ccmM and ccmN genes. Such mutants were shown to either contain aberrant carboxysomes or lack carboxysomes altogether (49, 131, 133).

Recent work suggests that carboxysomes, intracellular polyhedral bodies present in some autotrophic prokaryotes, play an important role in concentrating C02 for carboxysome* associated RubisCO (107, 131, 132). CHAPTER VI

~~Concluding Remarks~~

The Calvin reductive pentose phosphate cycle serves as the primary route to synthesize cell material from C02 in most photosynthetic organisms. Ribulosc 1,5- bisphosphate carboxylase/oxygenase (RubisCO) is the key enzyme of this pathway and catalyzes the actual fixation of C02 into organic carbon (179). The catalytic capacity of

~~RubisCO depends on its activation status, and an enzyme, RubisCO activase, was discovered to be involved in RubisCO activation in land plants and green algae (126).~~

The initial effort of this dissertation was devoted to exploring the existence of RubisCO activase in cyanobacteria, microbial counterparts of land plants and eukaryotic algae. By using a plant gene probe, the RubisCO activase gene ( rca) was found to be present in

~~Anabaena sp. strain CA and was subsequently isolated from a DNA library of this cyanobacterium. The deduced amino acid sequence of the Anabaena sp. strain CA~~

~~RubisCO activase shares greater than 50% identity with mature plant enzymes, and the putative ATP-binding domains and residues important for catalysis are conserved.~~

~~However, the Anabaena sequence shows distinctive features at the N and C termini.~~

~~When used as a probe to examine the distribution of rca in other cyanobacteria, the~~

~~Anabaena sp. strain CA rca gene hybridized to genomic DNA from all heterocystous~~

Anabaena and Nostoc strains examined but not to DNA from unicellular and 124 nonhcterocystous cyanobacteria. Similar results were obtained when the plant probe was employed under conditions of low stringency (Chapter II). Recently, RubisCO activase antigens were reported to be present in the cytosol of the unicellular cyanobacterium

Synechococcus sp. strain PCC 7942 (48). Although our Southern blot analysis results can not rule out the possibility that nonheterocystous cyanobacteria contain RubisCO activase, they certainly suggest that the latter type of RubisCO activase, if it exists, must be genetically distinct from the activase from plants and heterocystous cyanobacteria.

~~The location of rca was identified in three Anabaena species, strain CA, strain~~

PCC 7120, and A. variabilis. The Anabaena rca gene was uniformly situated downstream from the genes encoding the large and small subunits of RubisCO ( rbcL and rbcS). However, the region between rbcS and rca appeared to vary in size, and also in nucleotide sequence, since the two open reading frames (ORF1 and ORF2) encoded within the intergenic region of Anabaena sp. strain CA were not found in the intergenic region of the other two strains (Chapter II and III). Like other RubisCO-possessing bacteria, rbcL and rbcS are linked in Anabaena species. TheAnabaena rbc cluster also contains an open reading frame, rbcX , between rbcL and rbcS (93, Chapter III). A homologue of rbcX (of smaller size) was also found in the rbcL/rbdS intergenic region of Synechocystis sp. strain PCC 6803. Another unicellular cyanobacterium,

Synechococcus sp. strain a-1, might also contain rbcX between rbcL and rbcS if frameshifts occurred during sequencing (205, Chapter III). By contrast, Synechococcus sp. strain PCC 6301 does not have such an open reading frame in the intergenic region

(2, 165, Chapter III). In several photosynthetic and chemosynthetic bacteria, rbcL and 125 rbcS are closely associated with structural genes encoding other enzymes of the Calvin cycle, e.g., phosphoribulokinase (44, 61, 77, 112, 183). This is not the case in cyanobacteria, since studies with Anabaena sp. strain CA and Synechococcus sp. strain

~~PCC 7942 showed no indication that other Calvin cycle genes are present in the flanking regions of the rbc genes (49, 101, 158, Chapter V).~~

The main goal of this dissertation was to gain insights into the physiological role of RubisCO activase in Anabaena species. To achieve this objective, transcription of rca and rbc under various growth conditions was examined and a rca mutant was constructed, The expression of the Anabaena rca and rbc genes was coordinately, but differentially, regulated by light. During autotrophic growth of Anabaena cells, rca and rbc transcripts accumulated in the light and diminished in the dark. This light-dependent accumulation was controlled at the transcription level, since the stability of Anabaena sp. strain CA rca and rbc transcripts was diminished in the light. However, the rate of transcript accumulation and degradation for Anabaena sp. strain CA rca and rbc in response to variations in illumination were distinctive. The expression of the rca gene appeared to be much more sensitive than the rbc genes to light and dark transitions. In addition, light-dependent expression of these genes was not affected by the nitrogen source and the concentration of exogenous C02 supplied to the cells. When grown on fructose, rca and rbc transcripts accumulated in A. variabilis regardless of whether the cells were illuminated, but transcript levels were much lower in dark-grown cultures.

The expression of the rca and rbc genes in photoheterotrophic cultures of A. variabilis was similar to that in cultures grown with COj as the sole source of carbon (Chapter III). 126 The requirement for RubisCO activase in the light was further examined by measuring RubisCO activity of wild type and rca mutant strains of A. variabilis at different light intensities. In a 1 % C02-air atmosphere, inactivation of rca reduced the ability of A. variabilis to elevate RubisCO activity under high light (73 pmoles quanta m 1 s'1), but had little effect under low light (8 pmoles quanta nv2 s'1). For air-grown cultures, differences in the rates exhibited by the wild type and rca mutant to fully activate RubisCO during a whole-cell assay were enhanced by increases in light intensity.

The importance of RubisCO activase under high light conditions was also paralleled by the incapacity of the rca mutant to increase the growth rate after cells were transferred from low light to high light intensities. A high exogenous C02 concentration (1 %) was required to sustain a normal growth rate for the A. variabilis rca mutant. When grown in air-levels of C02, the rca mutant not only needed longer times to double in cell density but also exhibited greatly diminished RubisCO activity compared with the wild type strain. Since the results of Western blot analysis indicated that the wild type and mutant cells contained similar levels of RubisCO, the decline of carboxylation in the air cultures of the rca mutant was apparently due to diminished carbamylation (activation) of the enzyme (Chapter IV).

The high C02 requirement for growth and impaired light activation of RubisCO were also observed with activase-deflcient plants (110, 154). Plant RubisCO activase was suggested to modulate RubisCO activity by facilitating the release of metabolic inhibitors, such as ribulose 1,5-bisphosphate, xylulose 1,5-bisphosphate, ketoarabinitol

1,5-bisphosphate, and carboxyarabinitol 1-phosphate, from the active sites (143, 144, 127 194). However, our in vitro characterization studies with recombinant Anabaena sp. strain CA RubisCO showed that the Anabaena enzyme was only slightly inhibited by these phosphorylated compounds; indeed catalytic activity of the ER form of the enzyme rapidly recovered when MgJ+ and HCO/ were added to the reaction mixture. The specificity of Anabaena activase is thus obviously different from the plant enzyme

(Chapter IV). Thus far, the mechanism by which activase regulates RubisCO activation in Anabaena spp. is still not clear; but if metabolite inhibitors arc part of the regulation, one approach to gain more information might be to examine the activity of purified

~~Anabaena RubisCO after adding cell extracts prepared from wild type and rca mutant cells of A. variabilis grown under various conditions.~~

Although the rbcL-rbcS and rca genes arc linked and are in the same transcriptional orientation in Anabaena strains, hybridization of rbc and rca to distinct transcripts suggested that these genes are not cotranscribed, consistent with the results of primer extension and secondary structure analysis of the nucleotide sequence.

~~Expression of Anabaena sp. strain CA rca in Escherichia coli further suggested that the~~

~~154-bp 5' noncoding sequence was important for transcription of the rca gene (Chapter~~

III). Analysis of the regulatory elements within and flanking the 154 bp and their response to environmental changes certainly will help our understanding of the function of RubisCO activase. Transcription from ORF1 and ORF2 was not detected under the conditions examined, and the function of these putative genes remains unknown (Chapter

~~III). On the other hand, the product of rbcX might play some role in the assembly of~~

~~Anabaena RubisCO, since lack of expression of rbcX increased the dependence on 128 chaperonin proteins to produce properly assembled RubisCO holoenzymes in E. coli~~

~~(Chapter IV). LIST OF REFERENCES~~

~~1. Aitken A and Stanicr RY (1979) Characterization of pcptidoglycan from the cyanelles of Cyanophora paradoxa. J. Gen. Microbiol. 112: 219-223.~~

2. Amichay D, Levitz R and Gurevitz M (1993) Construction of a Synechocystis PCC6803 mutant suitable for the study of variant hcxadecameric ribulose bisphosphate carboxylase/oxygenase enzymes. Plant Mol. Biol. 23: 465-476.

~~3. Andcrsson I, Knight St Schneider G, Lindqvist Y, Lundqvist T, Branden C-I and Lorimer GH (1989) Crystal structure of the active site of ribulose-bisphosphate carboxylase. Nature 337: 229-234.~~

~~4. Andrews TJ (1988) Catalysis by cyanobacterial ribulose-bisphosphate carboxylase large subunits in the complete absence of smalt subunits. J. Biol. Chem. 263: 12213-12219.~~

~~5. Andrews TJ and Abel KM (1981) Kinetics and subunit interactions of ribulose bisphosphate carboxylase-oxygenase from the cyanobacterium Synechococcus sp. J. Biol. Chem. 256: 8445-8451.~~

6. Andrews TJ and Bailment B (1984) Active-site carbamate formation and reaction- intermediate-analog binding by ribulosebisphosphate carboxylase/oxygenase in the absence of its small subunits. Proc. Natl. Acad. Sci. USA 81: 3660-3664.

~~7. Andrews TJ and Kane HJ (1991) Pyruvate is a by-product of catalysis by ribulosebisphosphate carboxylase/oxygenase. J. Biol. Chem. 266: 9447-9452.~~

~~8. Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG, Smith JA and Struhl K (1989) Current protocols in molecular biology. Wiley, NY.~~

~~9. Badger MR, Price GD and Yu JW (1991) Selection and analysis of mutants of the COj-concentrating mechanism in cyanobacteria. Can. J. Bot. 69: 974-983.~~

~~10. Baldauf SL and Palmer JD (1990) Evolutionary transfer of the chloroplast tu/A gene to the nucleus. Nature 344: 262-265.~~

129 130 11. Barraclough R and Ellis RJ (1980) Protein synthesis in chloroplasts. IX. Assembly of newly-synthesized large subunits into ribulose bisphosphate carboxylase in isolated intact pea chloroplasts. Biochim. Biophys. Acta 608: 19- 31.

~~12. Beudeker RF and Tabita FR (1983) Control of photorespiratoiy glycolate metabolism in an oxygen-resistant mutant of Chlorella sorokiniana, J. Bacteriol. 155: 650-656.~~

~~13. Bimboim HC and Doly J (1979) A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Res. 7: 1513-1517.~~

14. Blake MS, Johnston KH, Russcl-Joncs GS and Gotschlich EC (1984) A rapid sensitive method for detection of alkaline- phosphatase-conjugated anti-antibody on Western blots. Anal. Biochem. 136: 175-179.

15. Boczar BA, Delaney TP and Cattolico RA (1989) Gene for the ribulose-1,5- bisphosphate carboxylase small subunit protein of the marine chromophytc Olisihodtscus luteus is similar to that of a chemoautotrophic bacterium. Proc. Natl. Acad. Sci. USA 86: 4996-4999.

16. Borbcly G and Simoncsits A (1981) 3'-terminal conserved loops of 16S rRNAs from the cyanobacterium Synechococcus PCC 6301 and maize chloroplast differ only in two bases. Biochem. Biophys. Res. Commun. 101: 846-852.

~~17. Boyer HB and Roulland-Sussaix D (1969) A complementation analysis of the restriction and modification of DNA in Escherichia coli. J. Mol. Biol. 41: 459-472.~~

~~18. Bryant DA Stirewalt VL (1990) The cyanelle genome of Cyanophora paradoxa encodes ribosomal proteins not encoded by the chloroplast genomes of higher plants. FEBS Lett. 259: 273-280.~~

~~19. Burke DH, Hearst JE and Sidow A (1993) Early evolution of photosynthesis: clues from nitrogenase and chlorophyll iron proteins. Proc. Natl. Acad. Sci. USA 90: 7134-7138,~~

~~20. Cai Y and Wolk CP (1990) Use of a conditionally lethal gene in Anabaena sp. strain PCC 7120 to select for double recombinants and to entrap insertion sequences. J. Bacteriol. 172: 3138-3145.~~

21. Campbell WJ and Ogren WL (1990) A novel role for light in the activation of ribulosebisphosphate carboxylase/oxygenase. Plant Physiol. 92:110-115. 131 22. Campbell WJ and Ogren WL (1990) Electron transport through photosystem I stimulates light activation of ribulose bisphosphate carboxylase/oxygenase (RubisCO) by RubisCO activase. Plant Physiol. 94:479-484.

23. Cardon ZG and Mott KA (1989) Evidence that ribulose 1,5-bisphosphate (RuBP) binds to inactive sites of RuBP carboxylase in vivo and an estimate of the rate constant for dissociation. Plant Physiol. 89: 1253-1257.

~~24. Castenholz RW (1970) Laboratory culture of thermophilic cyanophytcs. Schweiz Z Hydro!. 32: 538-551.~~

~~25. Chapman MS, Suh SW, Curmi PMG, Cascio D, Smith WW and Eiscnbcrg DS (1988) Tertiary structure of plant RuBisCO: domains and their contacts. Science 241: 71-74.~~

~~26. Chu DK and Bassham JA (1973) Activation and inhibition of ribulose 1,5- diphosphatc carboxylase by 6-phosphogluconatc. Plant Physiol. 52: 373-379.~~

~~27. Codd GA (1988) Carboxysomes and ribulose bisphosphate carboxylase/oxygenase. Adv. Microbiol. Physiol. 29: 115-164.~~

~~28. Codd CA and Stewart WDP (1977) Ribulose-l,5-diphosphate carboxylase in heterocysts and vegetative cells of Anabaena cylindrica. FEMS Microbiol. Lett. 2: 247-249.~~

29. Cozens AL and Walker JE (1987) The organization and sequence of the genes for ATP synthase subunits in the cyanobacterium Synechococcus 6301; support for an endosymbiotic origin of chloroplasts. J. Mol. Biol. 194: 359-383.

~~30. Curtis SE (1988) Light-energy transduction in photosynthesis: higher plants and bacterial models. In Stevens SE Jr and Bryant DA (eds). Am. Soc. Plant Physiol., Rockville.~~

31. Curtis SE and Haselkom R (1983) Isolation and sequence of the gene for the large subunit of ribulose 1,5-bisphosphate carboxylase from the cyanobacterium Anabaena 7120. Proc. Natl. Acad. Sci. USA 80: 1835-1839.

32. Curtis SE and Haselkom R (1984) Isolation, sequence, and expression of two members of the 32kd thylakoid membrane protein gene family from the cyanobacterium Anabaena 7120. Plant Mol. Biol. 3: 249-258.

33. Davis RW, Botstein D and Roth JR (eds.) (1980) Advanced bacterial genetics: a manual for genetic engineering. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.. 132 34. Destombc C and Douglas SE (1991) Rubisco spacer sequence divergence in the rhodophyte alga Gracilaria verrucosa and closely related species. Curr. Genet, 19: 395-398.

35. Ditta G, Schmidhauser T, Yakobson E, Lu P, Liang X-W, Finlay DR, Guincy D and Helcinski DR (1985) Plasmids related to the board host range vector, pRK290, useful for gene cloning and for monitoring gene expression. Plasmid 13: 149-153.

~~36. Douglas SE and Dumford DG (1989) The small subunit of ribulose-1,5- bisphosphate carboxylase is plastid-encoded in the chlorophyll c-containing alga Cryptomonas $. Plant Mol. Biol. 13: 13-20.~~

~~37. Duggan JX and Anderson, L.E. (1975) Light-rcgulation of enzyme activity in Anacystis nidulans (richt.). Planta 122: 293-297.~~

~~38. Edmondson DL, Badger MR and Andrews TJ (1990) A kinetic characterization of slow inactivation of ribulosebisphosphate carboxylase during catalysis. Plant Physiol. 93: 1376-1382.~~

39. Edmondson DL, Badger MR and Andrews TJ (1990) Slow inactivation of ribulosebisphosphate carboxylase during catalysis is not due to decarbamytalion of the catalytic site. Plant Physiol. 93: 1383-1389.

40. Edmondson DL, Badger MR and Andrews TJ (1990) Slow inactivation of ribulosebisphosphate carboxylase during catalysis is caused by accumulation of a slow tight-binding inhibitor at the catalytic site. Plant Physiol. 93: 1390-1397.

~~41. Edmondson DL, Kane HJ and Andrews TJ (1990) Substrate isomerization inhibits ribulosebisphosphate carboxylase-oxygenase during catalysis. FEBS Lett. 260:62- 66.~~

~~42. Elhai J and Wolk CP (1988) Conjugal transfer of DNA to cyanobacteria. Methods in Enzymology 167: 747-754.~~

~~43. Ellis RJ (1981) Chloroplast proteins: synthesis, transport, and assembly. Ann. Rev. Plant Physiol. 32: 111-139.~~

44. Falcone DL and Tabita FR (1993) Complementation analysis and regulation of CO; fixation gene expression in a ribulose 1,5-bisphosphate carboxylase- oxygenase deletion strain of Rhodospirillum rubrum. J. Bacteriol. 175: 5066- 5077. 133 43. Fay P (1992) Oxygen relations of nitrogen fixation in cyanobacteria. Microbiol. Rev. 56: 340-373.

46. Figurski D, Meyer R, Miller DS and Helinski DR (1976) Generation in vitro of deletions in the broad host range plasmid RK2 using phage Mu insertions and a restriction endonuclease. Gene 1: 107-119.

47. Fitchen JH, Knight S, Andcrsson I, Branden C-I and McIntosh L (1990) Residues in three conserved regions of the small subunit of ribulose-l,5-bisphosphatc carboxylase/oxygenase arc required for quaternary structure. Proc. Natl. Acad. Sci. USA 87: 5768-5772.

48. Fricdbcrg D, Jagcr KM, Kessel M, Silman NJ and Bergman B (1993) Rubisco but not Rubisco activase is clustered in the carboxysomes of the cyanobacterium Synechococcus sp. PCC 7942: Mud-induced carboxysomeless mutants. Mol. Microbiol. 9: 1193-1201.

49. Friedbcrg D, Kaplan A, Ariel R, Kessel M and Seijffers J (1989) The 5’-flanking region of the gene encoding the large subunit of ribulose-1,5 bisphosphate carboxylase/oxygenase is crucial for growth of the cyanobacterium Synechococcus sp. strain PCC 7942 at the level of C02 in air. J. Bacteriol. 171: 6069-6076.

~~50. Friedman AM, Long SR, Brown SE, Buikema WJ and Ausubel FM (1982) Construction of a broad host rang cosmid cloning vector and its use in the genetic analysis of Rhizobium mutants. Gene 18: 289-296,~~

51. Gantt JS, Baldauf SL, Calie PJ, Weeden NF and Palmer JD (1991) Transfer of rp)22 to the nucleus greatly preceded its loss from the chloroplast and involved the gain of an intron. EMBO J. 10: 3073-3078.

~~52. Gatenby AA (1984) The properties of the large subunit of maize ribulose bisphosphate carboxylase/oxygenase synthesized in Escherichia coli. Eur. J. Biochem. 144: 361-366.~~

~~53. Gay P, Le Coq D, Steinmetz M, Berkelman T and Kado Cl (1985) Positive selection procedure for the entrapment of insertion sequence elements im Gram- negative bacteria. J. Bacteriol. 164: 918-921.~~

54. Gay P, Le Coq D, Steinmetz M, Ferrari E and Hoch JA (1983) Cloning structural gene sacB, which codes for exoenzyme levansucrase of Bacillus subtilis'. expression of the gene in Escherichia coli. J. Bacteriol. 153: 1424-1431.

~~55. Georgiadis MM, Komiya H, Chakrabarti P, Woo D, Komuc JJ and Rees DC (1992) Crystallographic structure of the nitrogenase iron protein from Azotobacter 134 vinelandii. Science 257: 1653*1659.~~

56. Gibson JL (1994) Genetic analysis of CO; fixation genes. In: Blankenship RE, Madigan MT and Bauer CE (ed.) Anoxygenic photosynthetic bacteria, in press. Kluwer Acad. Pub., Dordrecht, The Netherlands.

57. Gibson JL, Chen C-H, Tower PA and Tabita FR (1990) The form II fructose 1,6-bisphosphatasc and phosphoribulokinasc genes form part of a large operon in Rhodobacter sphaeroides: primary structure and inscrtional mutagenesis analysis. Biochcmistiy 29: 8085*8093.

58. Gibson JL, Falcone DL and Tabita FR (1991) Nucleotide sequence, transcriptional analysis, and expression of genes encoded within the form I CO; fixation operon of Rhodobacter sphaeroides. J. Biol. Chem. 266: 14646*14653.

~~59. Gibson JL and Tabita FR (1977) Different molecular forms of d-ribuIosc-1,5- bisphosphate carboxylase from Rhodopseudomonas sphaeroides, J. Biol. Chem. 252: 943-949.~~

60. Gibson JL and Tabita FR (1986) Isolation of the Rftodopseudomonas sphaeroides form I ribulose 1,5-bisphosphate carboxylase/oxygenase large and small subunit genes and expression of the active hcxadecamcric enzyme in Escherichia coli. Gene 44: 271-278.

~~61. Gibson JL and Tabita FR (1988) Localization and mapping of C02 fixation genes within two gene clusters in Rhodobacter sphaeroides. J. Bacteriol. 170: 2153- 2158.~~

~~62. Gibson JL and Tabita FR (1993) Nucleotide sequence and functional analysis of CbbR, a positive regulator of the Calvin cycle operons of Rhodobacter sphaeroides. J. Bacteriol. 175: 5778-5784.~~

~~63. Giovannoni SJ, Turner S, Olsen GJ, Bams S, Lane DJ and Pace NR (1988) Evolutionary relationships among cyanobacteria and green chloroplasts. J. Bacteriol. 170: 3584-3592.~~

~~64. Golden SS, Brusslan J and Haselkom R (1987) Genetic engineering of the cyanobacteria! chromosome. Methods in Enzymology 153: 215-231,~~

65. Golden JW, Carrasco CD, Mulligan ME, Schneider GJ and Haselkom R (1988) Deletion of a 55-kilobase-pair DNA element from the chromosome during heterocyst differentiation of Anabaena sp. strain PCC 7120. J. Bacteriol. 170: 5034-5041. 135 66. Golden JW, Whorff LL and Wiest DR (1991) Independent regulation of nifHDK operon transcription and DNA rearrangement during heterocyst differentiation in the cyanobacterium Anabaena sp. strain PCC 7120. J. Bacteriol. 173: 7098*7105.

67. Goloubinoff P, Christeller JT, Gatcnby AA and Lorimer GH (1989) Reconstitution of active dimeric ribulose bisphosphate carboxylase from an unfolded state depends on two chapcronin proteins and Mg-ATP. Nature 342: 884*889.

~~68. Goloubinoff P, Gatcnby AA and Lorimer GH (1989) GroE heat-shock proteins promote assembly of foreign prokaiyotic ribulose bisphosphate carboxylase oligomers in Escherichia coli. Nature 337: 44-47.~~

~~69. Gotto JW, Tabita FR and Van Baalcn C (1979) Isolation and characterization of rapidly-growing, marine, nitrogen-fixing strains of blue-green algae. Arch. Microbiol. 121: 155-159.~~

70. Gurevitz M, Somerville CR and McIntosh L (1985) Pathway of assembly of ribulose bisphosphate carboxylase/oxygenase from Anabaena 7120 expressed in Escherichia coli. Proc. Natl. Acad. Sci. USA 82: 6546-6550.

~~71. Harley CB and Reynolds RP (1987) Analysis of E. coli promoter sequences. Nuct. Acids Res. 15: 2343-2361.~~

72. Hartman FC and Harpel M (1993) Chemical and genetic probes of the active site of D-ribulose-1,5-bisphosphate carboxylase/oxygenase: a retrospective based on the three-dimensional structure. Adv. Enzymol. 67: 1-75.

~~73. Haury JF and Spiller H (1981) Fructose uptake and influence on growth of and nitrogen fixation by Anabaena variabilis. J. Bacteriol. 147: 227-235.~~

74. Hemmingsen SM, Woolford C, van der Vies SM, Tilly K, Dennis DT, Georgopoulos CP, Hendrix RW and Ellis RJ (1988) Homologous plant and bacterial proteins chaperone oligomeric protein assembly. Nature 333: 330-334.

~~75. Herrero A, Elhai J, Hohn B and Wolk CP (1984) Infrequent cleavage of cloned Anabaena variabilis DNA by restriction endonucleases from A. variabilis. J. Bacteriol. 160: 781-784.~~

76. Higgins CF, Hiles ID, Salmond GPC, Gill DR, Downie JA, Evans IJ, Holland IB, Gray L, Buckel SD, Bell AW and Hermodson MA (1986) A family of related ATP-binding subunits coupled to many distinct biological processes in bacteria. Nature 323: 448-450. 136 77. Husemann M, Klintworth R, Biittchcr V, Salnikow J, Wcisscnbom C and Bowien B (1988) Chromosomally and plasmid-encoded gene clusters for C02 fixation (cfx genes) in Alcaligenes eutrophus. Mol. Gen. Genet. 214: 112-120.

~~78. Hwang S-R and Tabita FR (1989) Cloning and expression of the chloroplast- cncoded rbcL and rbcS genes of the marine diatom Cylindrotheca sp. strain Nl. Plant Mol. Biol. 13: 69-79.~~

79. Hwang S-R and Tabita FR (1991) Cotranscription, deduced primaiy structure, and expression of the chloroplast-encoded rbcL and rbcS genes of the marine diatom Cylindrotheca sp. strain Nl. J. Biol. Chcm. 266: 6271-6279.

~~80. Jensen RG and Bahr JT (1977) Ribulose 1,5-bisphosphate carboxylasc-oxygenasc. Annu. Rev. Plant Physiol. 28: 379-400.~~

~~81. Jordan DB and Chollet R (1983) Inhibition of ribulose bisphosphatc carboxylase by substrate ribulose 1,5-bisphosphatc. J. Biol. Chem. 258: 13752-13758.~~

~~82. Jordan DB, Chollet R and Ogrcn WL (1983) Binding of phosphorylatcd effectors by active and inactive forms of ribulose- 1,5-bisphosphate carboxylase. Biochemistry 22: 3410-3418.~~

~~83. Jordan DB and Ogren WL (1981) Species variation in the specificity of ribulose bisphosphate carboxylase/oxygenase. Nature 291: 513-515.~~

~~84. Keen NT, Tamaki S, Kobayashi D and Trollinger D (1988) Improved broad-host- range plasmids for DNA cloning in Gram-negative bacteria. Gene 70: 191-197.~~

~~85. Klintworth R, Huseman M, Salnikow J and Bowien B (1989) Chromosomal and plasmid locations for phosphoribulokinase genes in Alcaligenes eutrophus. J Bacteriol 2: 954-956.~~

~~86. Knauf VC and Nester EW (1982) Wide host range cloning vectors: a cosmid clone bank of an Agrobacterium Ti plasmid. Plasmid 8: 45-54.~~

~~87. Knight S, Andersson I, and Branden C-I (1990) Ciystallographic analysis of ribulose 1,5-bisphosphate carboxylase from spinach at 2.4 A resolution. J. Mol. Biol. 215: 113-160.~~

~~88. Kossmann J, Klintworth R and Bowien B (1989) Sequence analysis of the chromosomal and plasmid genes encoding phosphoribulokinase from Alcaligenes eutrophus. Gene 85: 247-252. 137~~

89. Kusano T, Takcshima T, Inoue C and Sugawara K (1991) Evidence for two sets of structural genes coding for ribulose bisphosphate carboxylase in ThiobaciUus ferrooxidans. J. Bacteriol. 173: 7313-7323.

~~90. Laing WA, and Christeller JT (1976) A model for the kinetics of activation and catalysis of ribulose 1,5-bisphosphate carboxylase. Biochem. J. 159: 563-570.~~

91. Lan Y and Mott KA (1991) Determination of apparent Km values for ribulose 1,5-bisphosphate carboxylase/oxygenase (RubisCO) activase using the spcctrophotomctric assay of RubisCO activity. Plant Physiol. 95: 604-609.

92. Lang JD, and Haselkom R (1989) Isolation, sequence and transcription of the gene encoding the photosystcm II chlorophyll-binding protein, CP-47, in the cyanobacterium Anabaena 7120. Plant Mol. Biol. 13: 441-456.

~~93. Larimer FW and Soper TS (1993) Overproduction of Anabaena 7120 ribulosc- bisphosphatc carboxylase/oxygenase in Escherichia coli. Gene 126: 85-92.~~

94. Lee B, Bcrka RM and Tabita FR (1991) Mutations in the small subunit of cyanobacteria ribulose-bisphosphatc carboxylase/oxygenase that modulate interactions with large subunits. J. Biol. Chem. 266: 7417-7422.

95. Lee B, Read BA and Tabita FR (1991) Catalytic properties of recombinant octameric, hexadecameric, and heterologouscyanobacterial/bactcrial ribulose-1,5- bisphosphate carboxylase/oxygenase. Arch Biochem. Biophys. 291: 263-269.

96. Lee B and Tabita FR (1990) Purification of recombinant ribulose-1,5- bisphosphate carboxylase/oxygenase large subunits suitable for reconstitution and assembly of active LgS* enzyme. Biochemistry 29: 9352-9357.

~~97. Lewin RA and Withers NW (1975) Extraordinary pigment composition of a prokaryotic alga. Nature 256: 735-737.~~

98. Li L-A, Gibson JL and Tabita FR (1993) The RubisCO activase (rca) gene is located downstream from rbcS in Anabaena sp. strain CA and is detected in other Anabaena/Nostoc strains. Plant Mol. Biol. 21: 753-764.

~~99. Li L-A and Tabita FR (1994) Transcription control of ribulose bisphosphate carboxylase/oxygenase activase and adjacent genes in Anabaena species. J. Bacteriol. 176. In press.~~

100. Li L-A, and Tabita FR Unpublished data. 138 101. Lieman-Hurwitz J, Schwarz R, Martinez F, Maor Z, Reinhold L and Kaplan A (1991) Molecular analysis of high C02 requiring mutants; involvement of genes in the region of rbc , including rbcS, in the ability of cyanobacteria to grow under low COj. Can. J. Bot. 69: 945-950.

~~102. Lloyd JC, Horsncll PR, Dyer TA, and Raines CA (1991) Structure and sequence of wheat phosphoribulokinase gene. Plant Mol. Biol. 17: 167-168.~~

~~103. Lorimer GH (1981) Ribulosebisphosphate carboxylase: amino acid sequence of a peptide bearing the activator carbon dioxide. Biochemistry 20: 1236-1240.~~

104. Lorimer GH, Badger MR and Andrews TJ (1976) The activation of ribulose-1,5- bisphosphate carboxylase by carbon dioxide and magnesium ions. Equilibria, kinetics, a suggested mechanism, and physiological implications. Biochcmistty 15: 529-536.

~~105. Lugtcnbcrg B, Mcijers J, Peters R, Van der Hoek P and Van Alpen L (1975) Electrophoretic resolution of the ‘major outer membrane protein* of E. coli into bands. FEBS Lett. 58: 254-258.~~

~~106. Maniatis T, Fritsch EF and Sambrook J (1982) Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.~~

~~107. Marcus Y, Berry JA and Pierce J (1992) Photosynthesis and photorespiration in a mutant of the cyanobacterium Synechocystis PCC 6803 lacking carboxysomes. Planta 187: 511-516.~~

~~108. Margulis L (1981) Symbiosis in cell evolution W. H. Freeman and Co., San Francisco.~~

~~109. Marsden WJN and Codd GA (1984) Purification and molecular and catalytic properties of phosphoribulokinase from the cyanobacterium Chlorogloeopsis fritschii. J. Gen. Microbiol. 130: 999-1006.~~

110. Mate CJ, Hudson GS, von Caemmerer S, Evans JR and Andrews TJ (1993) Reduction of ribulose bisphosphate carboxylase activase levels in tobacco (Nicotiana tabacum) by antisense RNA reduces ribulose bisphosphate carboxylase carbamylation and impairs photosynthesis. Plant Physiol. 102: 1119-1128.

111. McCam DF, Whitaker RA, Alam J, Vrba JM and Curtis SE (1988) Genes encoding the alpha, gamma, delta, and four F0 subunits of ATP synthase constitute an operon in the cyanobacterium Anabaena sp. strain PCC 7120. J. Bacteriol. 170: 3448-3458. 139 112. Meijcr WG, Amberg AC, Enequist HG, Tcrpstra P, Lidstrom ME and Dijkhuizen L (1991) Identification and organization of carbon dioxide fixation genes in Xanthobacter JJavus H4-14. Mol. Gen. Genet. 225: 320-330,

~~113. Milanez S and Mural RJ (1988) Cloning and sequencing of cDNA encoding the mature form of phosphoribulokinase from spinach. Gene 66: 55-63.~~

~~114. Milanez S, Mural RJ and Hartman FC (1991) Role of cysteinyl residues of phosphoribulokinase as examined by site-directed mutagenesis. J. Biol. Chem. 266: 10694-10699.~~

~~115. Miziorko HM and Lorimer GH (1983) Ribulose-1,5-bisphosphate carboxylase- oxygcnase. Ann. Rev. Biochem. 52: 507-535.~~

~~116. Mizobata T, Akiyama Y, Ito K, Yumoto N and Kawata Y (1992) Effects of the chaperonin GroE on the folding of tryptophanase from Escherichia coli. J. Biol. Chem. 267: 17773-17779.~~

~~117. Mulligan B, Schultes N, Chen L and Bogorad L (1984) Nucleotide sequence of a multiple-copy gene for the B protein of photosystem II of a cyanobacterium. Proc. Natl. Acad. Sci. USA 81: 2693-2697.~~

~~118. Nargang F, McIntosh L and Somerville C (1984) Nucleotide sequence of the ribulosebisphosphate carboxylase gene from RItodospirillum rubrum. Mol. Gen. Genet. 193: 220-224.~~

119. Newman SM, Derocher J and Cattolico RA (1989) Analysis of chromophytic and rhodophytic ribulose-1,5-bisphosphate carboxylase indicates extensive structural and functional similarities among evolutionarily diverse algae. Plant Physiol. 91: 939-946.

120. Nierzwicki-Bauer SA, Curtis SE and Haselkom R (1984) Cotranscription of genes encoding the small and large subunits of ribulose-l,5-bisphosphate carboxylase in the cyanobacterium Anabaena 7120, Proc. Natl. Acad. Sci. USA 81: 5961- 5965.

121. Parry MAJ, Keys AJ, Foyer CH, Furbank RT and Walker DA (1988) Regulation of ribulose-1,5-bisphosphate carboxylase activity by the activase system in lysed spinach chloroplasts. Plant Physiol. 87: 558-561.

122. Paul K, Morell MK and Andrews TJ (1991) Mutations in the small subunit of ribulosebisphosphate carboxylase affect subunit binding and catalysis. Biochemistry 30: 10019-10026. 140 123. Pierce J, Carlson TJ and Williams JGK (1989) A cyanobacteria! mutant requiring the expression of ribulose bisphosphate carboxylase from a photosynthetic anaerobe. Proc. Natl. Acad. Sci. USA 86: 5753-5757.

~~124. Porter MA and Hartman FC (1988) Catalytic nonessentiality of an active-site cysteinyl residue of phosphoribulokinase. J. Biol. Chem. 263: 14846-14849.~~

~~125. Porter MA, Stringer CD and Hartman FC (1988) Characterization of the regulatory thiorcdoxin site of phosphoribulokinase. J. Biol. Chem. 263: 123-129.~~

~~126. Portis AR Jr (1990) Rubisco activase. Biochim. Biophys. Acta 1015: 15-28.~~

~~127. Portis AR Jr (1992) Regulation of ribulose 1,5-bisphosphate carboxylase/oxygenase activity. Annu. Rev. Plant Physiol. Plant Mol. Biol. 43: 415-437.~~

128. Portis AR Jr, Salvucci ME and Ogren WL (1986) Activation of ribulose bisphosphate carboxylase/oxygenase at physiological C02 and ribulose bisphosphate concentrations of RubisCO activase. Plant Physiol. 82: 967-971.

~~129. Prentki P and Krisch HM (1984) In vitro inscrtional mutagenesis with a selectable DNA fragment. Gene 29: 303-313.~~

~~130. Price GD and Badger MR (1989) Expression of human carbonic anhydrase in the cyanobacterium Synechococcus PCC7942 creates a high C02-requiring phenotype. Plant Physiol. 91: 505-513.~~

~~131. Price GD and Badger MR (1991) Evidence for the role of carboxysomes in the cyanobacterial C02-coneentration mechanism. Can. J. Bot. 69: 963-973.~~

~~132. Price GD, Coleman JR and Badger MR (1992) Association of carbonic anhydrase activity with carboxysomes isolated from the cyanobacterium Synechococcus PCC7942, Plant Physiol. 100: 784-793.~~

133. Price GD, Howitt SM, Harrison K and Badger MR (1993) Analysis of a genomic DNA region from the cyanobacterium Synechococcus sp. strain PCC7942 involved in carboxysome assembly and function. J. Bacteriol. 175: 2871-2879.

~~134. Pridmore RD (1987) New and versatile cloning vectors with a kanamycin- resistance marker. Gene 56: 309-312.~~

135. Ramos JL, Madueno F and Guerrero MC (1985) Regulation of nitrogenase levels in Anabaena sp. ATCC 33047 and other filamentous cyanobacteria. Arch. Microbiol. 141: 105-111. 141 136. Read BA and Tabita FR (1992) Amino acid substitutions in the small subunit of ribulose- 1,5-bisphosphate carboxylase/oxygenase that influence catalytic activity of the holoenzyme. Biochemistry 31: 519-525.

~~137. Read BA and Tabita FR (1992) A hybrid ribulosebisphosphate carboxylase/oxygenase enzyme exhibiting a substantial increase in substrate specificity factor. Biochemistry 31: 5553-5560.~~

138. Read BA and Tabita FR (1994) High substrate specificity factor ribulose bisphosphate carboxylase/oxygenase from eukaryotic marine algae and properties of recombinant cyanobactcrial Rubisco containing "algal" residue modifications. Arch. Biochem. Biophys. 312: 210-218.

~~139. Rcinhold L, Kosloff R and Kaplan A (1991) A model for inorganic carbon fluxes and photosynthesis in cyanobacteria! carboxysomes. Can. J. Bot. 69: 984-988.~~

~~140. Rice D, Mazur BJ and Haselkom R (1982) Isolation and physical mapping of nitrogen fixation genes from the cyanobacterium Anabaena 7120. J. Biol. Chem. 257: 13157-13163.~~

~~141. Rippka R, Derucllcs J, Watcrbury JB, Herdman M and Stainer RY (1979) Generic assignments, strain histories and properties of pure cultures of cyanobacteria. J. Gen. Microbiol. Ill: 1-61,~~

142. Robinson SP and Portis AR Jr (1988) Involvement of stromal ATP in the light activation of ribulose-1,5-bisphosphate carboxylase/oxygenase in intact isolated chloroplasts. Plant Physiol. 86: 293-298.

143. Robinson SP and Portis AR Jr (1988) Release of the nocturnal inhibitor, carboxyarabinitol-l-phosphate, from ribulose bisphosphate carboxylase/oxygenase by RubisCO activase. FEBS Lett. 223: 413-416.

144. Robinson SP and Portis AR Jr (1989) Ribulose-1,5-bisphosphate carboxylase/oxygenase activase protein prevents the in vitro decline in activity of ribulose-l,5-bisphosphate carboxylase/oxygenase. Plant Physiol. 90: 968-971.

~~145. Robinson SP and Portis AR Jr (1989) Adenosine triphosphate hydrolysis by purified RubisCO activase. Arch. Biochem. Biophys. 268: 93-99.~~

~~146. Robinson SP, Streusand VJ, Chatfield JM and Portis AR Jr (1988) Purification and assay of Rubisco activase from leaves. Plant Physiol. 88: 1008-1014.~~

~~147. Roesler KR and Ogren WL (1988) Nucleotide sequence of spinach cDNA encoding phosphoribulokinase. Nucl. Acids Res. 16: 7192. 142~~

~~148. Roesler KR and Ogrcn WL (1990) Chlamydomonas reinhardtii phosphoribulokinase. Plant Physiol. 93: 188-193.~~

149. Roesler KR and Ogrcn WL (1990) Primary structure of Chlamydomonas reinhardtii ribulose 1,5-bisphosphate carboxylase/oxygcnase activase and evidence for a single polypeptide. Plant Physiol. 94: 1837-1841.

~~150. Roy H and Cannon S (1988) Ribulose bisphosphate carboxylase assembly: what is the role of the large subunit binding protein? Trends Biochem. Sci. 13: 163- 165.~~

~~151. Rundle SJ and Zielinski RE (1991) Organization and expression of two tandemly oriented genes encoding ribulosbisphosphatc carboxylase/oxygcnase activase in barley. J. Biol. Chem. 266: 4677-4685.~~

152. Rundle SJ and Zielinski RE (1991) Alterations in barley ribulose-1,5-bisphosphate carboxylase/oxygenase activase gene expression during development and in response to illumination. J. Biol. Chem. 266: 14802-14807.

~~153. Salvucci ME, Portis AR Jr and Ogren WE (1985) A soluble chloroplast protein catalyzes ribulosebisphosphate carboxylase/oxygcnase activation in vivo. Photosyn. Res. 7: 193-200.~~

~~154. Salvucci ME, Portis AR Jr and Ogrcn WL (1986) Light and C02 response of ribulose-1,5-bisphosphate carboxylase/oxygenase activation inArabidopsis leaves. Plant Physiol. 80: 655-659.~~

~~155. Salvucci ME, Wemeke JM, Ogrcn WL and Portis AR Jr (1987) Purification and species distribution of Rubisco activase. Plant Physiol. 84: 930-936.~~

~~156. Sanger F, Nicklen S and Coulsen AR (1977) DNA sequencing with chain- terminating inhibitors. Proc. Natl. Acad. Sci. USA 74: 5463-5467.~~

~~157. Schneider GJ, Lang JD and Haselkom R (1991) Promoter recognition by the RNA polymerase from vegetative cells of the cyanobacterium Anabaena 7120. Gene 105: 51-60.~~

158. Schwarz R, Lieman-Hurwitz J, Hassidim M and Kaplan A (1992) Phenotypic complementation of high CCVrequiring mutants of the cyanobacterium Synechococcus sp. strain PCC 7942 by inosine 5’-monophosphate. Plant Physiol. 100: 1987-1993.

159. Schweizer HP (1992) Allelic exchange in Pseudomonas aeruginosa using novel ColEl-type vectors and a family of cassettes containing a portable oriT and the 143 counter-selectable Bacillus sublilis sacB marker. Mol. Microbiol. 6:1195-1204.

~~160. Seewaldt E and Stackcbrandt E (1982) Partial sequence of 16S ribosomal RNA and the phytogeny of Prochloron. Nature 295: 618-620.~~

~~161. Scrra JL, Llama MJ, Rowell P and Stewart WDP (1989) Purification and characterization of phosphoribulokinase from the Nj-fixing cyanobacterium Anabaena cylindrica. Plant Science 59: 1-9.~~

~~162. Servaitcs JC (1990) Inhibitionof ribulose 1,5-bisphosphate carboxylase/oxygcnase by 2-carboxyarabinitol-l-phosphatc. Plant Physiol. 92: 867-870,~~

163. Shcn JB, Orozco EM Jr and Ogrcn WL (1991) Expression of the two isoforms of spinach ribulose 1,5-biphosphate carboxylase activase and essentiality of the conserved lysine in the consensus nucleotide-binding domain. J. Biol. Chem. 266: 8963-8968.

164. Shine J and Dalgamo L (1974) The 3’-tcrminal sequence of Escherichia coli 16S ribosomal RNA: complementarity to nonsense triplets and ribosome binding sites. Proc. Natl. Acad. Sci. USA 71: 1342-1346.

165. Shinozaki K and Sugiura M (1983) The gene for the small subunit of ribulose- 1,5-phosphate carboxylase/oxygenase is located close to the gene for the large subunit in the cyanobacterium Anacystis nidulans 6301. Nucl. Acids Res, 11: 6957-6954.

166. Shinozaki K and Sugiura M (1985) Genes from the large and small subunits of ribulose-1,5-bisphosphate carboxylase/oxygcnase constitute a single operon in a cyanobacterium Anacystis nidulans 6301. Mol. Gen. Genet. 200: 27-32.

~~167. Shively JM and English RS (1991) The carboxysome, a prokaryotic organelle: a mini-review. Can. J. Bot. 69: 957-962.~~

~~168. Smith SM and Ellis RJ (1981) Light-stimulated accumulation of transcripts of nuclear and chloroplast genes for ribulosebisphosphate carboxylase. J. Mol. Appl. Genet. 1: 127-137.~~

169. Smrcka AV, Bohnert HJ and Jensen RG (1991) Modulation of the tight binding of carboxyarabinitol 1,5-bisphosphate to the large subunit of ribulose 1,5- bisphosphate carboxylase/oxygenase. Arch. Biochem. Biophys. 286: 14-19.

170. Somerville CR, Portis AR Jr and Ogren WL (1982) A mutant of Arabidopsis thaliana which lacks activation of RuBP carboxylase in vivo. Plant Physiol. 70: 381-387. 144 171. Staccy G, Van Baalcn C and Tabita FR (1977) Isolation and characterization of a marine Anabaena sp. capable of rapid growth on molecular nitrogen. Arch. Microbiol. 114: 197-201.

172. Stein JL, Haygood M and Felbeck H (1990) Nucleotide sequence and expression of a deep-sea ribulose bisphosphate carboxylase gene cloned from a chemoautotrophic bacterial symbiont. Proc.Natl. Acad. Sci. USA 87:8850-8854.

~~173. Stoscheck CM (1990) Quantitation of protein. Methods in Enzymology 182: 50- OS.~~

~~174. Streusand VJ and Portis AR Jr (1987) Rubisco activase mediates ATP-dcpcndcnt activation of ribulose bisphosphate carboxylase. Plant Physiol. 85: 152-154.~~

~~175. Su X and Bogorad L (1991) A residue substitution in phosphoribulokinase of Synechocystis PCC 6803 renders the mutant light sensitive. J. Biol. Chem. 266: 23698-23705.~~

~~176. Suwanto A and Kaplan S (1989) Physical and genetic mapping of the Rhodobacter sphaeroides genome: presence of two unique circular chromosomes. J. Bacteriol. 171: 5850-5859.~~

~~177. Tabita FR (1980) Pyridine nucleotide control and subunit structure of phosphoribulokinase from photosynthetic bacteria. J. Bacteriol. 143: 1275-1280.~~

~~178. Tabita FR (1987) Carbon dioxide fixation and its regulation in cyanobacteria. In Fay P and Van Baalen C (eds), The Cyanobacteria, pp. 96-117. Elsevier, Amsterdam.~~

~~179. Tabita FR (1988) Molecular and cellular regulation of autotrophic carbon dioxide fixation in microorganisms. Microbiol. Rev. 52: 155-189.~~

180. Tabita FR (1994) The biochemistry and molecular regulation of carbon dioxide metabolism in cyanobacteria. In: Bryant DA (ed.) The Molecular Biology of Cyanobacteria, pp. 299-329. Kluwer Acad. Pub., Dordrecht, The Netherlands.

181. Tabita FR (1994) The biochemistry and metabolic regulation of carbon metabolism and CO* fixation in purple bacteria. In: Blankenship RE, Madigan MT and Bauer CE (ed.) Anoxygenic photosynthetic bacteria, in press. Kluwer Acad. Pub., Dordrecht, The Netherlands.

182. Tabita FR, Caruso P and W Whitman (1978) Facile assay of enzymes unique to the Calvin Cycle in intact cells with special reference to ribulose 1,5-bisphosphate carboxylase. Anal. Biochem. 84: 462-472. 145 183. Tabita FR, Gibson JL, Falcone DL, Wang X, Li L-A, Reed BA, Terlesky KC and Paoli GC (1993) Current studies on the molecular biology and biochemistry of C02 fixation in phototrophic bacteria. In: Murrell C and Kelly PD (eds.) Microbial Growth on Ct Compounds pp.469-479. Intercept Scientific Publication, Andover, UK.

~~184. Tabita FR and McFadden BA (1974) D-ribulose 1,5-diphosphate carboxylase from Rltodospirillum rubrum. J. Biol. Chem. 249: 3459-3464.~~

185. Tabita FR and Small CL (1985) Expression and assembly of active cyanobactcrial ribulose-1,5-bisphosphate carboxylase/oxygenase in Escherichia coli containing stoichiometric amounts of large and small subunits. Proc. Natl. Acad. Sci. USA 82:6100-6103.

~~186. Thiel T and Wolk CP (1987) Conjugal transfer of plasmids to cyanobacteria. Methods in Enzymology 153: 232243.~~

~~187. Thome SW, Newcomb EH and Osmond CB (1977) Identification of chlorophyll b in extracts of prokaryotic algae by fluorescence spectroscopy. Proc. Natl. Acad. Sci. USA 74: 575-578.~~

~~188. Valentin K and Zetsche K (1989) The genes of both subunits of ribulose-1,5- bisphosphate carboxylase constitute an operon on the plastid of a red alga. Curr. Genet. 16: 203-209.~~

~~189. Van Baalen C (1962) Studies on marine blue-green algae. Bot. Mar. 4: 129-139.~~

190. van den Bergh ERE, Dijkhuizen L and Meijer WG (1993) CbbR, a LysR-type transcriptional activator, is required for expression of the autotrophic C02 fixation enzymes of Xanthobacter Jlavtts. J. Bacteriol. 175: 6097-6104.

191. Viale AM, Kobayashi H and Akazawa T (1989) Expressed genes for plant-type ribulose 1,5-bisphosphate carboxylase/oxygenase in the photosynthetic bacterium Chromatium vinosum, which possess two complete sets of the genes. J.Bacteriol. 171: 2391-2400.

192. Walker JE, Saraste M, Runswick MJ and Gay NJ (1982) Distantly related sequences in the a- and fl-subunits of ATP synthase, myosin, kinases and other ATP-rcquiring enzymes and a common nucleotide binding fold. EMBO J. 1: 945- 951,

193. Wang X, Falcone DL and Tabita FR (1993) Reductive pentose phosphate- independent C02 fixation in RJiodobacter sphaeroides and evidence that RubisCO activity serves to maintain the redox balance of the cell. J. Bacteriol. 175: 3372- 146

~~3379.~~

194. Wang ZY and Portis AR, Jr (1992) Dissociation of ribulose-1,5-bisphosphatc bound to ribulose-1,5-bisphosphate carboxylase/oxygcnase and its enhancement by ribulose-1,5-bisphosphate carboxylase/oxygcnase activase-mcdiatcd hydrolysis of ATP. Plant Physiol. 99: 1348-1353.

~~195. Wasmann CC, Loffelhardt W and Bohnert HJ (1987) Cyanelles: organization and molecular biology. In Fay P and Van Baalcn C (eds), The cyanobacteria pp. 303-324. Elsevier, Amsterdam, N.Y.~~

~~196. Weaver KE and Tabita FR (1985) Complementation of a Rhodopseudomonas sphaeroides ribulose bisphosphate carboxylasc-oxygenase regulatory mutant from a genomic library. J. Bacteriol. 164: 147-154.~~

197. Wemeke JM, Chatficld JM and Ogren WL (1988) Catalysis of ribulosebisphosphate carboxylase/oxygcnase activation by the product of a RubisCO activase cDNA clone expressed in Escherichia coii. Plant Physiol. 87: 917-920.

~~198. Wemeke JM, Zielinski RE and Ogren WL (1988) Structure and expression of spinach leaf cDNA encoding ribulosebisphosphate carboxylase/oxygenase activase. Proc. Natl. Acad. Sci. USA 85: 787-791,~~

~~199. Whitman W and Tabita FR (1976) Inhibition of ribulose 1,5-bisphosphate carboxylase by pyridoxal 5-phosphate. Biochem. Biophys. Res. Commun. 71: 1034-1039.~~

~~200. Whitton BA (1972) The biology of the blue-green algae. In Carr NG, Whitton BA (eds) pp. 415-433. Blackwell Scientific, Oxford.~~

201. Winder TL, Anderson JC amd Spalding MH (1992) Translational regulation of the large and small subunits of ribulose bisphosphate carboxylase/oxygenase during induction of the C02-concentrating mechanism in Chlamydomonas reinhardtii. Plant physiol. 98: 1409-1414.

~~202. Windhdvel U and Bowien B (1990) On the operon structure of the cfit gene clusters in Alcaligenes eutrophus. Arch. Microbiol. 154: 85-91.~~

~~203. Windhdvel U and Bowien B (1991) Identification of cfitR, an activator gene of autotrophic C02 fixation in Alcaligenes eutrophus. Mol. Microbiol, 5:2695-2705.~~

204. Wolk CP (1973) Physiology and cytological chemistry of blue-green algae. Bact. Rev. 37: 32-101, 147 265. Yaguchi T, Chung SY, Igarashi Y and Kodama T (1993) Cloning, sequence and ovcrexpression of the thermophilic cyanobacterium gene for the ribulose-1,5- bisphosphate carboxylase/oxygcnase. J. Fcrmentat. Biocng. 75: 1-8.

~~206. Yakota A (1991) Carboxylation and detoxification of xylulose bisphosphate by spinach ribulose bisphosphate carboxylase/oxygcnase. Plant Cell Physiol. 32: 755-762.~~

~~207. Yanisch-Peron C, Vieira J and Messing J (1985) Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mpl8 and pUC19 vectors. Gene 33: 103-119.~~

~~208. Zhu G and Jensen RG (1991) Xylulose 1,5-bisphosphate synthesized by ribulose 1,5-bisphosphate carboxylase/oxygcnase during catalysis binds to dccarbmylatcd enzyme. Plant Physiol. 97: 1348-1353.~~

~~209. Zhu G and Jensen RG (1991) Failover of ribulose 1,5-bisphosphate carboxylase/oxygcnase activity. Plant Physiol. 97: 1354-1358.~~

~~210. Zielinski RG, Wemeke JM and Jenkins ME (1989) Coordinate expression of RubisCO activase and RubisCO during barley leaf cell development. Plant Physiol. 90: 516-521.~~