sign in sign up
<<
Home , Chloroplast DNA, DNA replication, Genome, HeLa, Origin of replication, Photoautotrophism

Order N u m b e r 9201787

Manipulation of the

Ye, Jingsong, Ph.D.

The Ohio State University, 1991

UMI 300 N. Zeeb Rd. Ann Arbor, MI 48106 MANIPULATION OF THE CHLOROPLAST GENOME

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of the Ohio State University

By

Jingsong Ye, B.S., M.S.

*****

The Ohio State University 1991

Dissertation Committee: Approved by R. Sayre V . Raghavan F. Sack Advisor Department of D. Bisaro To my Father and Mother To my wife Hua and my daughter Qian-qian

• * 11 ACKNOWLEDGMENTS I would like to express my sincere appreciation and gratitude to all those involved in the completion of this dissertation: Dr. Richard T. Sayre, my advisor, for his continually valuable advice and guidance during the investigation and preparation of the dissertation. His gentle nature, kindness and personal concerns greatly helped me complete the study and adapt to the in the U.S.A. Dr. V. Raghavan, Dr. F. Sack, Dr. D. Bisaro, Dr. T. Sims and Dr. R. Hangarter for their advice and suggestions concerning my Ph.D. project and my career. Ms. Susan Ketchner, Ms. Robin Roffey, Ms. Elizabeth Wrobel-Boerner, Mr. John Okuley and Mr. Reed Clark for their guidance and assistance in laboratory work and for their kindness which made my life at The Ohio State University enjoyable and memorable. Finally, my wife, Yan Hua and my daughter, Qian-qian, for their endless love, encouragement and sacrifice during my study for Ph.D. VITA

May 4, 1957 Born in Nanjing, P.R. China 1982 B.S., Wuhan University, Hubei, P.R. China 1986 M.S. , University of Illinois at Urbana-Champaign 1986-Present Graduate Student Teaching Assistant, Ohio State Plant Biology Program, Columbus, Ohio

PUBLICATIONS Ye J, Hauptmann RM, Smith AG, Widholm JM (1987) Selection of a Nicotiana plumbaainifolia universal hybridizer and in intergeneric somatic hybrid formation. Mol Gen Genet 208:474-480 Ye, J, Sayre RT (1990) Reduction of chloroplast DNA content in Solanum nigrum suspension cells by treatment with chloroplast DNA synthesis inhibitors. Plant Physiology 94:1477-1483

FIELDS OF STUDY Major Field: Plant Biology Studies in development of a universal somatic hybrid selection system for plant protoplast fusion under Dr. Jack M. Widholm Studies in manipulation of chloroplast genome of Solanum nigrum and Chlamydomonas in vivo under Dr. Richard T. Sayre TABLE 07 CONTENTS

ACKNOWLEDGMENTS ...... iii VITA ...... iv LIST OF TABLES...... viii LIST OF FIGURES...... ix ABBREVIATIONS ...... xi CHAPTER PAGE I. INTRODUCTION ...... 1 1. 1 INTRODUCTION...... 1 1.2 REGULATION of CHLOROPLAST DNA SYNTHESIS ...... 2 1.2.1 Chloroplast Genome ...... 2 1.2.2 DNA Copy Number...... 4 1.2.3 Chloroplast DNA Replication ...... 7 1.2.4 Inhibitors of Chloroplast DNA Replication . 9 1.3 CHLOROPLAST TRANSFORMATION ...... 11 1.3.1 Introduction of DNA into the Nucleus 12 1.3.2 Introduction of Foreign Plastid ...... 13 1.3.3 Autonomously Replicating ...... 14 1.3.4 Introduction of Foreign into ...... 15 1.3.5 Problems and Prospects...... 18 1.4 TRANSFORMATION OF CHLAMYDOMONAS CHLOROPLASTS _____ 19 1.4.1 Chiamydomonas is a Unique Transformation System...... 19 1.4.2 Delivery of DNA to Chloroplasts...... 20 1.4.3 Selectable Marker for Chloroplast Transformation ...... 25 1.5 RECA ...... 28 1.5.1 General Characteristic ...... 28 1.5.2 Role in General Recombination...... 32 1.5.3 Role in DNA Re p a i r ...... 40 1.5.4 RecA Protein Analogues...... 44

v II.REDUCTION OF CHLOROPLAST DNA CONTENT IN SOLANUM NIGRUM SUSPENSION CELLS BY TREATMENT WITH CHLOROPLAST DNA SYNTHESIS INHIBITORS ...... 46 2.1 INTRODUCTION...... 46 2.2 MATERIALS AND METHODS...... 48 2.2.1 Plant Material and Growth Conditions ... 48 2.2.2 Isolation of Protoplasts...... 48 2.2.3 Treatment of Suspension Cells with Chloroplast DNA Synthesis Inhibitors ... 49 2.2.4 Analyses of Chloroplast and Nuclear DNA Content...... 50 2.2.5 Quantification of Total Nucleic by Fluorescent DNA Binding D y e s ...... 52 2.3 RESULTS...... 53 2.3.1 Effects of DNA Synthesis Inhibitors on Growth and Viability...... 53 2.3.2 Effects of DNA Synthesis Inhibitors on Nuclear and Chloroplast DNA Content ..... 57 2.4 DISCUSSION...... 66 III. CHLOROPLAST TRANSFORMATION AND ...... 71 3.1 INTRODUCTION...... 71 3.2 MATERIALS AND METHODS...... 74 3.2.1 Plasmids and Chlamvdomonas Strains ...... 74 3.2.2 Reduction of Chloroplast DNA Content in Chlamvdomonas ...... 75 3.2.3 Ch1amydomonas Transformation...... 76 3.2.4 of from Suspension Cells and Leaves...... 77 3.2.5 Transformation of Solanum niorum Cells ... 77 3.2.6 Southern Hybridization Analyses ...... 78 3.2.7 Preparation of Probe .... 79 3.2.8 DNA Hybridizations...... 80 3.2.9 Chloroplast Mutagenesis ...... 81 3.3 RESULTS ...... 81 3.3.1 Regulation of Chloroplast DNA Content in Chlamvdomonas...... 81 3.3.2 Transformation Frequency in Chlamvdomonas 84 3.3.3 Regeneration of Plants from Suspension and Leaf Cells ...... 85 3.3.4 Chloroplast Transformation and Mutagenesis...... 87 3.4 DISCUSSION...... 95 IV. CHLOROPLAST RECA-LIKE PROTEIN IN CHLAMYDOMONAS ___ 104 4.1 INTRODUCTION...... 104 4.2 MATERIALS AND METHODS...... 106

Vi 4.2.1 Chlamvdomonas and ...... 106 4.2.2 Insolation of Chloroplast and Western Blotting a..**...... 10*7 4.2.3 Treatment of Chlamvdomonas Cells with ...... 108 4.2.4 Labeling Plasmid DNA with Tritiated ...... 109 4.2.5 DNA Renaturation As s a y ...... 110 4.2.6 Purification of RecA-like Protein ...... Ill 4.2.7 Preparation of pX13 s s D N A ...... 112 4.2.8 Preparation of [MP]-dsDNA...... 113 4.2.9 Assay of Three-stranded D N A ...... 114 4.3 RESULTS...... 115 4.3.1 Identification of "RecA-like" Protein in the Chloroplast of Chiamydomonas ...... 115 4.3.2 Renaturation of D N A ...... 120 4.3.3 Determination of Optimal Conditions for Chloroplast Protein Catalyzed DNA Renaturation Reaction...... 123 4.3.4 Partial Purification of RecA-like Protein...... 140 4.3.5 Formation of Three-Stranded D N A ...... 14 6 4.4 DISCUSSION...... 155 V. SUMMARY...... 161 5.1 REDUCTION cpDNA CONTENT BY TREATMENT WITH cpDNA SYNTHESIS INHIBITORS ...... 161 5.2 TRANSFORMATION OF S. NIGRUM AND MUTAGENIC EFFECTS OF CpDNA SYNTHESIS INHIBITORS...... 163 5.3 REGULATION OF cpDNA CONTENT AND TRANSFORMATION FREQUENCY IN CHLAMYDOMONAS ...... 164 5.4 RecA-LIKE PROTEIN IN CHLAMYDOMONAS CHLOROPLASTS ...... 165 5.5 PROSPECTS...... 168 REFERENCES...... 170

vii LIST OF TABLES TABLE 1. Effects of DNA Synthesis Inhibitors on the DNA Content and Fresh Weight of £. nigrum Suspension Cells (8 day Culture) ...... 58 2. Effects of DNA Synthesis Inhibitors on the DNA Content and Fresh Weight of £. nigrum Suspension Cells (2X4 day culture) ...... 59 3. Quantification of Total DNA in Protoplasts Labeled with the Fluorescent DNA Binding Dye, DAPI ...... 63 4. Effects of Chloroplast DNA Synthesis Inhibitors on DNA Content of Protoplasts Isolated from 2 X 4 day Suspension Cultures of S. nigrum as Determined by Dot Blot Hybridizations...... 65 5. Effects of cpDNA Synthesis Inhibitors on DNA Content of Chlamydomonas as Determined by Dot Blot Hybridizations ...... 83 6. Increase of the Transformation Frequency for Chlamydomonas Chloroplast psbA Deletion ...... 86 7. Effect of BAP and IAA on Green Shoot Formation from Callus...... 88 8. Mutagenic Effect of Novobiocin on Chloroplast DNA ... 96 9. Ammonium Sulfate Fractionation ...... 141 10. Cellulose Purification...... 142 11. DEAE Sephadex Purification...... 143 LIST or FIGURES FIGURE 1. Schematic of a Microprojectile Delivery System .... 21 2. Reactions Promoted by recA Protein...... 34 3. Reaction Pathway for the Transfer of a DNA Strand from a Linear Duplex to a Circular s s D N A ...... 39 4. Model of the SOS Regulatory System...... 42 5. of S. nigrum Cultures ...... 54 6. Effects of Inhibitors on Growth of Cell Suspensions o f S ■ ...... a...... 55 7. Dot Blot of a Dilution Series of DNA Extracted from S. nigrum Protoplasts ...... 63 8. Restriction maps of Tag 1 Fragment of psbA in &. Hvbridus. £. nigrum and Possible Transformants ...... 90 9. Southern Blot Analyses of Selected Atrazine Resistant Plants...... 91 10. Stringency Washes for Selected Atrazine Resistant Plants...... 93 11. Western Blot Analysis of Polypeptides from Ch1amvdomonas Cells probed with the £. coli RecA ...... 117 12. Effects of Novobiocin on the Levels of RecA-like Protein from Chloroplast of Ch1amvdomonas ...... 118 13. Annealing of Single Strand DNA by RecA-like Protein in Stromal Extraction from Chlamvdomonas Chloroplast...... 121 14. Effect of pH on RecA-like Protein Catalyzed DNA Renaturation...... 124

ix 15. Effect of Temperature on RecA-like Protein Catalyzed DNA Renaturation...... 126 16. Effect of ATP Concentration on RecA-like Protein Catalyzed DNA Renaturation...... 129 17. Kinetics of ATP Dependent Renaturation Reaction Catalyzed by RecA-like Protein ...... 131 18. Effect of NaCl Concentration on RecA-like Protein Catalyzed DNA Renaturation...... 132 19. Effect of KC1 Concentration on RecA-like Protein Catalyzed DNA Renaturation...... 134 20. Effect of MgCl, Concentration on RecA-like Protein Catalyzed DNA Renaturation...... 136 21. Effect of CaCl2 Concentration on RecA-like Protein Catalyzed DNA Renaturation ...... 138 22. Annealing of Single Strand DNA by RecA-like Protein in Crude Extract, Ammonium Sulfate Fraction and DEAE Fraction...... 144 23. Purification of RecA-like Protein in Chlamvdomonas Chloroplasts ...... 147 24. Protein Separation by Sodium Deodecyl Sulfate (SDS)-Polyacrylamide Gel Electrophoresis ...... 149 25. Formation of Joint Molecules by £. coli R e c A ..... 151 26. Formation of Joint Molecules by RecA-like Protein...... 153

x ABBRBVIATIONN

ADP, ; ARC, autonomous replication in Chlamydomonas; ARS, autonomously replicating sequence; ATP, ; BAP, 6-benzylaminopurine; , chloramphenicol acetyl transferase; cpDNA, chloroplast DNA; 2,4-D, 2,4-dichlorophenoxy acetic ; D-loops, displacement loops; dsDNA, double-stranded DNA; DAPI, 4,6-diamido-2- phenylindole; DTT, dithiothreitol; EDTA, ethylene diaminetetraacetic acid; Fudr, 5-fluordeoxyuridine; HS medium, high salt medium; IAA, indoleacetic acid; kbp, kilobase pair; LB medium, Luria-Bertani medium; HS medium, Nurashige-Skoog medium; NADH, nicotinamide dinucleotide (reduced form); PEG, polyethylene glycol; rbcL, ribulose-1,5- biphosphate carboxylase large subunit; SDS, sodium dodecyl sulfate; SSC, 0.15 H NaCl, 0.017 M Na citrate (pH 7.0); ssDNA, single-stranded DNA; SSPE, 0.15 M NaCl, 0.01 M NaH2P04, 0.001 M EDTA (pH 7.4); ssRubisco, subunit of ribulose-1,5 bisphosphate carboxylase; STE, 10 mM Tris-HCl (pH 8.0), 100 mM NaCl,l mM EDTA; TAP, Tris-acetate-phosphate; TBE, 89 mM Tris- borate, 89 mM boric acid, 2 mM EDTA; Tm, melting temperature; TSA, 20 mM Tris-HCl (pH 7.4), 150 mM NaCl, 1.5 g/1 sodium azide; TY medium, 1.6% tryptone, 1% extract, 0.5% NaCl.

xi CHAPTER I INTRODUCTION

1.1 INTRODUCTION Plastids are a form of that are unique to plants and certain protista. Plastids exist in a number of different forms with different functions, but the chloroplast is the best characterized of all plastids. The discovery, at the beginning of this century, of non-Mendelian inherited giving rise to altered chloroplast phenotypes suggested the existence of a separate genetic system in chloroplasts. Various mutations resulting in defects have been shown to be inherited maternally through cytoplasm indicating that chloroplast genes play an essential role in the normal growth of photosynthetic . Since the demonstration of the chloroplast genome over 20 years ago, intensive studies on the structure and organization of chloroplast genes, and on the identity, sequence and expression of chloroplast genes have been made (Whitfield and Bottomley, 1983; Well, 1987; Mullet, 1988; Shimada and Sugiura, 1991). Several important systems are localized in the chloroplast including: the photosynthetic electron transport

1 2 and carbon fixation pathways, the sulfur and nitrogen reduction pathways, lipid and biosynthetic pathways as well as a variety of secondary metabolic pathways. Obviously, the manipulation of many of these biochemical pathways, ultimately to alter the quality of plants, has significant agronomic as well as basic research value. The alteration of chloroplast DNA encoded proteins through genetic manipulation, however, requires a large amount of background information on the location and expression of genes for plastid proteins, and the means of introducing genes into a selected genome of the plant for proper expression. In this chapter the background information relevant to the development of chloroplast genetic manipulation will be reviewed and discussed. Specific emphasis will be placed on the regulation of chloroplast DNA (cpDNA) synthesis, chloroplast transformation and cpdna recombination.

1.2 REGULATION OF CHEOROPIA8T DNA SYNTHESIS 1.2.1 Chloroplast Genome Almost all chloroplast that have been examined by electron microscopy and restriction mapping exist as a single, more or less homogeneous size class of circular molecules. Host chloroplast DNAs fall into a rather restricted size range of between 120 kilobase pairs (kbp) and 160 kbp. Only among green does one encounter a wide range of cpDNA sizes from 85 kbp to 292 kbp. However, the chloroplast of 3 vascular plants and most algae are quite similar in general structure and organization. One of the outstanding features of a typical cpDNA molecule is the presence of a large . The repeated sequences are separated by a large and a small single-copy region (Sugita et al., 1984). The complete sequences of cpDNA from the liverwort, Marchantia polvmorpha (Ohyama et al., 1986) and from tobacco (Shinozaki et al., 1986) have been determined and provide a major source of information on the organization of the chloroplast genome. Chloroplast encoded genes identified to date can be divided into two groups. One group encodes proteins and RNA molecules involved in the chloroplast and processes. For example, all tRNA species involved in chloroplast protein synthesis are believed to be coded in chloroplast genome. Thirty different tRNA genes have been identified in the cpDNA genome (Kato et al., 1981, 1985; Deno et al., 1982; Deno and Sugiyura, 1983, 1984; Ohme et al., 1984, 1985; Sugita et al., 1984, 1985; Yamada et al., 1986; Wakasugi et al., 1986). Chloroplast DNA also encodes a complete set of ribosomal (23s, 16s, 5s and 4.5s). In general, the chloroplast rRNA genes are organized into a large transcription unit in the inverted repeats and thus are present in two copies per genome (Strittmatter et al. 1985). The largest class of chloroplast genes are those that proteins of the photosynthetic apparatus. The large subunit of ribulose-1,5-bisphosphate carboxylase/oxygenase 4 (rbcL), the major soluble protein of the chloroplast, is chloroplast encoded. At least twenty-two genes encode proteins of the membrane complex including I, photosystem II, cytochrome b^/f complexes and the ATP synthase. In addition, six open reading frames have been identified which have sequence with mitochondrial genes encoding complete subunits of the respiratory chain NADH dehydrogenase. Several of the remaining unidentified reading frames are conserved between diverse species which suggests that they may also encode functional plastid polypeptides.

1.2.2 Plastid DNA Copy Number Many copies of plastid DNA are normally present in each cell and in each plastid as well. Each plastid contains ten to several hundred copies of the chloroplast genome and there may be as many as 20-50 chloroplasts per cell, thus each cell has a high number of cpDNA. mesophyll cells, three to eight cm above the leaf base, contains 8,000 to 12,000 copies of plastid DNA which is distributed in 60 plastids (Baumgartner et al. , 1989). Such high plastid DNA copy numbers have also been reported in leaf cells of spinach (up to 12,000; Lawrence and Possingham, 1986) and (10,000; Lamppa and Bendich, 1979). Furthermore, the number of chloroplast genomes per organelle and the number of chloroplasts per cell in higher 5 plants may vary depending on age, tissue type and developmental stage of the cell (Boffey et al., 1979; Boffey and Leech, 1982; Lawrence and Possingham, 1986; Miyammra et al., 1986; Baumgartner et al., 1989). For example, plastid DNA copy number has been reported to increase 7.5 fold in leaf primordial cells of wheat grasses during germination and development of leaf mesophyll cells. This increase in plastid DNA copy number per cell, which occurs during mesophyll cell development, is due to an increase in the number of cpDNA copies per plastid and an increase in plastid number (Boffey et al., 1979; Boffey and Leech 1982). It has been reported that colorless, dark grown H* tabacum cells contain approximately 3,300-4,800 plastid genome copies per cell, whereas light grown green cells contain about 9,500-12,000 chloroplast genomes per cell (Cannon et al., 1985). In various etiolated tissues the plastid DNA copy number is also subject to change as demonstrated for pea, mung-bean (Thompson et al., 1983) and spinach (Aguettaz et al., 1987). In cultured tobacco cells, preferential synthesis of plastid DNA and increased replication of plastids has been observed following medium renewal (Yasuda et al., 1988). The copy number of plastid DNA per cell determined by fluorescence microspectroscopy using DNA binding dyes increased 11-fold within one day of culture to reach 11,000 copies/cell, then decreased gradually to 1,000 copies/cell after one week of culture. Since active plastid DNA synthesis occurs under 6 heterotrophic conditions in darkness, it is possible that the copy number of plastid DNA in cultured cells is influenced more strongly by the culture stages than by the light conditions. Although these observations show that plastid DNA copy number is variable, the significance of these variations with respect to plastid and regulation is not clear. By comparison of the levels of total run-on transcription products as a function of plastid DNA levels from dark-grown and illuminated cotyledons and young and mature leaves, Deng and Gruissem (1987) observed a decrease in plastid DNA during the 24 hour greening period with a concomitant 3-fold increase in overall transcriptional activity. During leaf maturation, however, they found an increase in chloroplast DNA with 5-fold decrease in overall transcriptional activity. These results suggest that plastid DNA and transcription levels are inversely correlated. Similarly, mutant Chiamydomonas cells with only one-fourth the wild-type amount of cpDNA per cell have nearly normal photosynthetic capacity (Hosier et al., 1989). Thus one can predict that cpDNA copy number does not significantly influence protein synthesis and may not substantially influence normal rates of photosynthetic growth. However, there are several conflicting reports regarding these conclusions. Aguettaz et al. (1987) measured the relative variation of the plastid DNA content per spinach cell, the cellular levels of chloroplast rRNA transcripts and 7 rbcL transcripts. They found that the variation of RNA levels was directly correlated with the changes in cellular plastid DNA cellular levels. These results suggest that plastid DNA plays a central role in the regulation of plastid transcript levels. A direct relationship between decreased gene dosage and increased levels of rbcL mRNA has been demonstrated during differentiation of etioplasts to chloroplasts in pea (Inamine et al. , 1985; Sasaki et al., 1986).

1.2.3 Chloroplast DNA Replication The chloroplast is a semiautonomous, self-duplicating organelle with its own hereditary apparatus. The number of chloroplast and chloroplast genomes per cell may vary in response to changing metabolic needs of the cell. Therefore, it is interesting to know how cells synthesize cpDNA and how cpDNA synthesis is regulated. To date, most information about chloroplast DNA replication mechanisms has been obtained from electron microscopic studies, sequence analyses and 4D vitro replication assays. In the lower eukaryotic, unicellular alga, Eucrlena gracilis, the origin of chloroplast DNA replication has been mapped by electron microscopic studies to a 5-6 kbp sequence upstream of the 5' end of the supplementary 16s rRNA gene (Ravel-Chapuis et al., 1982; Koller and Delius, 1982). This DNA fragment contains a polymorphic region (Schlunegger et 8 al., 1983) which is composed of multiple tandem repeats. This region and its immediate vicinity are extremely A+T rich and have the capacity to form multiple stem-loop structures (Schlunegger and Stutz, 1984). In the algae Ch1amvdomonas reinhardtii. two putative replication origins containing displacement loops (D-loops) were found 16.5 kb upstream of the 5* end of the 16s rRNA gene. Replication initiated with the formation of the D-loop results in the synthesis of one daughter strand 400 bp in length (Waddell et al., 1984). Like DNA replication in , chloroplast DNA synthesis proceeds in both directions. Using an independent approach, Rochaix et al. (1984) isolated autonomously replicating sequences (ARC sequence) in Ch1amvdomonas cpDNA by transformation and selection for plasmids that would replicate autonomously in £. reinhardtii (ARC plasmids). one such plasmid, ARC1 of cpDNA origins, hybridized to a restriction fragment that contains the D-loop region cloned by Wang et al.(1984). Although ARC elements have been physically mapped and shown to be distinct from the origin of DNA replication (Vallet et al., 1984) both sites lie in close proximity to each other (Vallet and Rochaix, 1985). Wu et al. (1986) used an DNA replication system to study replication functions using cloned chloroplast DNA fragments containing the D-loop and flanking sequence. They found that the DNA fragments were active as templates for DNA synthesis in crude algal extracts that support DNA synthesis. 9 In higher plants, electron microscopic examination of pea and corn cpDNA has demonstrated the presence of Cairns type and rolling circle type structures during cpDNA replication, suggesting that cpDNA from higher plants replicates by both the Cairns and the rolling circle mechanism (Kolodner and Tewari, 1975). Based on structural homology with the putative cpDNA replication origin of Euolena gracilis, a potential replication origin in Petunia hvbrida cpDNA was proposed to be located in the single copy region close to the inverted repeat (de Hass et al., 1986). When a chimeric plasmid containing a DNA fragment with a potential chloroplast replication origin was used as a template for DNA replication in a Petunia hybrida chloroplast lysate system, specific initiation of DNA synthesis was observed (de Hass et al., 1987).

1.2.4 Inhibitors of Chloroplast DNA Replication Since the replication of chloroplast DNA is separated from nuclear DNA by time and space it is possible to specifically inhibit chloroplast DNA synthesis without affecting nuclear DNA replication. It has been reported that 5-fluorodeoxyuridine (Fudr) selectively inhibits chloroplast DNA replication in Chlamvdomonas (Wurts et al., 1977; Hurts et al., 1979) with little effect on the growth of the . Fudr (1 mM) reduces the amount of cpDNA 6 to 7 fold. In addition, Fudr treatment increases the number of maternally inherited mutations by 20 to 50 fold if cells are kept in 10 stationary phase for 24 hours. It Is postulated that Fudr acts Indirectly as a by causing starvation in the chloroplast only, leading to repression of cpDNA synthesis (Wurtz et al., 1979). Chloroplast replication in Chlamvdomonas has also been reported to be sensitive to , a DNA intercalating dye (Flechtner and Sager, 1973). In light-grown cells, ethidium bromide (10 ug/ml) inhibits the replication of cpDNA while permitting replication of nuclear DNA. Pre­ existing chloroplast DNA is partially degraded by this treatment. Under heterotrophic conditions, nuclear DNA replication is also inhibited by ethidium bromide. However, short-term exposure to ethidium bromide is reversible provided the drug is removed within 12 hours. Similar levels of ethidium bromide selectively reduce cpDNA content in by 70% while the effect on total cellular DNA is negligible (Heinhorst et al.1986) , an inhibitor of DNA gyrase (Gellert et al. 1977), also inhibits DNA replication in procaryotes. The unicellular alga Eualena gracilis can be irreversibly bleached when grown in the presence of nalidixic acid. This phenomenon is characterized by a severe loss of cpDNA with no apparent effect on (Pienkos et al., 1974). Similar results have been observed in tobacco (Heinhorst et al., 1985, 1986). These findings suggest that plastid DNA and nuclear DNA synthesis are not tightly coupled in cultured tobacco cells. 11 Lam and Chua (1987) have found that pea chloroplasts contain a gyrase-like enzyme co-existing with a I. Like the bacterial enzyme, the pea chloroplast gyrase is sensitive to novobiocin. With use of this inhibitor, they demonstrated that when relaxed templates are used, the chloroplast gyrase acts as a positive modulator of gene activity in vitro. Their results suggest that changes in template topology may be a mechanism by which chloroplast genes are differentially regulated and that chloroplast DNA gyrase and topoisomerase I are key for this mode of regulation in vivo.

1.3 CHLOROPLAST TRANSFORMATION Genes encoding plastid proteins are located either on the plastid genome or the nuclear genome, in which case the proteins are synthesized in the cytoplasm and targeted to the plastids. Therefore, several possible approaches must be considered for transformation of plastid encoded proteins in plants: (1) insert the gene of interest into the nuclear genome with sequences coding for nuclear transcription control signals and a plastid transit (2) introduce foreign chloroplasts containing the gene of interest into the host cell (3) introduce foreign cpDNA sequences into plasmids that replicate independently of cpDNA and (4) insert the gene directly into the chloroplast genome for expression by the chloroplast genetic apparatus. The following section will discuss these methods and emphasis will be placed on the 12 higher plants.

1.3.1 Introduction of DNA into the Nucleus One approach to manipulate chloroplast proteins is the introduction of chloroplast genes into the nuclear genome using Aarobacterium--derived vectors. Basically, the chloroplast transit peptide sequence of a nuclear encoded protein which is targeted to the chloroplast (such as small subunit of Rubisco) is fused to the 5' coding sequence of the gene of interest and introduced into the plant using a Ti plasmid vector. The hybrid protein is synthesized in the cytoplasm and is targeted to the chloroplast due to the presence of the amino terminal transit peptide sequence. After the hybrid protein crosses the chloroplast envelope, the transit sequence is cleaved by a specific protease. Both neomycin phosphotransferase and the chloroplast psbA atrazine resistance gene product have been successfully imported into chloroplasts using this approach (Van den Broek et al., 198 5; Schreier et al., 1985). Effectively, this is a protein mediated phenotypic transformation of the chloroplast and is not a true genetic transformation. This approach has certain limitations in contrast to a direct chloroplast DNA mediated transformation. Obviously, the temporal and quantitative control of gene expression in the nucleus and chloroplast may differ significantly. In addition, differences in copy number of the transforming gene (1-10 copies per cell) versus the 13 larger numbers of chloroplast genome copies (100-10,000 copies) may be a limitation. Furthermore, gene products which are normally encoded by the chloroplast genome but which are imported into the chloroplast as fusion products (e.g. psbA atrazine resistance gene products) must compete with the wild type chloroplast encoded copy for assembly into protein complexes and phenotypic expression. Finally, the inheritance of nuclear encoded and chloroplast targeted gene products may be complex due to segregation during sexual transmission.

1.3.2 Introduction of Foreign Plastids Protoplast fusion is the process by which the nuclei and the cytoplasm (including ) from two cells can be combined into one cell. By using a chloroplast selectable marker, a one-step transfer of chloroplasts has been achieved (Bellinad et al., 1978; Medgyesy et al., 1980; Henczel et al., 1981; Medgyesy et al., 1985). Metabolic complimentation, however, aided the recovery of interspecific fusion products which survived and formed calli. For example, Sidorov et al. (1981) used a system for the transfer of chloroplasts which did not make use of selectable genetic markers. In this system, protoplasts of Nicotiana tabacum were irradiated to inactivate nuclei and fused with iodoacetate-treated Nicotiana plumbaqanifolia protoplasts lacking the ability to divide. Callus formation was only possible from interspecific fusion products. Thus, irradiated J*. tabacum chloroplasts had an 14 improved chance of dominating the heterokaryon-derived cells. However, the resulting products from this chloroplast transfer system contained not only chloroplasts from both parents but a mixture of nuclei and other organelles as well. Incorporation of isolated organelles into plant cells permits transfer of a single type of organelle. There have been a number of reports of the successful uptake of isolated chloroplasts by plant protoplasts (Bonnett and Eriksson, 1974; Bonnett, 1976; Giles et al., 1980), but stable incorporation of the chloroplasts and subsequent regeneration of transformed plants have not been achieved.

1.3.3 Autonomously Replicating Plasmids Another possible mechanism for chloroplast transformation is the introduction of autonomously replicating plasmids. To construct such a plasmid, a sequence which will act as an in the chloroplast must be included. Unfortunately, as discussed above, a higher plant chloroplast origin of replication has not been sufficiently characterized. A number of studies have attempted to identify chloroplast origins of replication by their ability to promote autonomous replication in yeast. Autonomous replication sequences (ARS) which will replicate in yeast have been located and characterized from Petunia (de Hass et al., 1986) and tobacco cpDNA (Uchimiya et al., 1983; Ohtani et al., 1984) . However, it has been demonstrated that these ARS as 15 well as sequences promoting autonomous replication (ARC) in Chlamvdomonas are distinct from the origin of chloroplast DNA replication. In addition, no sequences that promote autonomous replication in yeast have been found in chloroplast DNA (Gold et al., 1987). Therefore, further studies on the origin of cpDNA replication are necessary for transformation of chloroplasts using autonomous replicating plasmids.

1.3.4 Introduction of Foreign Genes into Chloroplasts Another mechanism to transform the chloroplast is to integrate foreign DNA sequences into the chloroplast genome so that the foreign sequence will be replicated and inherited together with native chloroplast genes. One successful transformation of cpDNA has been achieved using tumefaciens T-DNA constructs (De Block et al., 1985). The T-DNA construct used for chloroplast transformation consisted of genes encoding nopaline synthase and the coding sequences of chloramphenicol acetyltransferase (CAT), aminoglycoside phosphotransferase type II (aphll) and B-lactamase. The construct was introduced into tobacco leaf protoplasts by co-cultivation with Aj. tumefaeiens containing the hybrid Ti plasmid. Chloramphenicol-resistant (Cmr) calli were selected on media containing 10 ug/ml chloramphenicol. Both CAT activity and aminoglycoside phosphotransferase activity were localized in the chloroplast by cell fractionation studies on leaves of Cmr plants. The results of 16 reciprocal crosses with wild type Nicotiana tabacum and the results of Southern blot hybridization analysis indicated the introduced genes were maternally inherited. However, CAT activity was lost if transformed plants were grown in the absence of chloramphenicol, suggesting that the CAT gene was not stably maintained in the chloroplast DNA. A number of issues were raised by this reported transformation of tobacco chloroplasts with the CAT gene, including the mechanism of DNA uptake by the chloroplast, the site of integration of the CAT gene, the fate of the gene constructs and the site and size of the transcripts of the CAT gene. Unfortunately, there have been no further reports of chloroplast transformation using Ti plasmid derived vectors. There are several other possible approaches for achieving chloroplast transformation. Microinjection is one of the most precise techniques for delivery of into specific intracellular compartments. This technique offers the great advantage that individual protoplasts containing chloroplasts may by transformed at high frequency. However, microinjection has not been demonstrated to be successful for chloroplast DNA transformation. One problem is that the number of chloroplasts per cell requiring injection is not clear since little is known of the fate of chloroplasts and cpDNA during the various stages of plant regeneration from protoplasts. Furthermore, the method requires considerable skill to carry out and requires some mechanism for integration 17 of foreign DNA into the cpDNA. The electroporation method is based on the use of short electrical impulses of high field strength that increase the permeability of the protoplast membrane (Zimmermann and Vienken, 1982). To use this method for chloroplast transformation, one must consider the possibility that chloroplasts possess a DNA uptake system, although nothing is known about such a DNA uptake system in chloroplasts. The reported transformation of chloroplasts by Agrobacterium tumefaciens (De Block et al., 1985) suggests that chloroplast may be able to take up exogenous DNA. Chloroplast DNA recombination in higher plants has also been demonstrated by selecting regenerated plants from protoplast fusions of cells containing different chloroplast encoded resistance markers (Medgesy et al., 1985). However, the observed frequency of chloroplast recombination events was very low. This is probably due to the large copy number of chloroplast genomes per cell and the attendant segregation. Another approach for delivering DNA to chloroplasts is the particle bombardment method, which was developed by Sanford et al. (1987). The advantage of this method is that with each bombardment thousands of tungsten particles are accelerated at the same time thereby delivering DNA into many cells or organs simultaneously. Since the tungsten projectiles carrying DNA can penetrate the cell wall of intact plant 18 cells, this transformation method overcomes the difficulties associated with working with protoplasts. Another advantage of this method is that neither cell type, size, shape nor species should significantly alter its effectiveness. By using the bombardment method, Maliga and his coworkers have reported genetic transformation of plastid genome in Nicotiana tabacum suggesting a possibility of experimental modification of higher plant plastid genome (Svab et al., 1990).

1.3.5 Problems and Prospects It appears from the above discussion that there is no easy way to transform chloroplasts. Aarobacterium-Ti plasmid mediated gene transformation does not appear to be an efficient and reproducible method for chloroplast transformation and other methods associated with the use of protoplasts, such as protoplast fusion and electroporation, are limited by the availability of plant regeneration from protoplasts. Furthermore, there is no information on whether chloroplasts have a DNA uptake and integration system. The particle bombardment method appears to be a more promising method to introduce foreign DNA into chloroplasts due to the universality of its application. However, even if this method is successfully developed, further questions concerning the integration, maintenance, replication and expression of foreign DNA in the chloroplasts must be addressed. Therefore, more studies are needed to provide the knowledge of 19 recombination, replication and transcription of cpDNA required for chloroplast transformation. The possibility of engineering a DNA uptake system into the chloroplast envelope may be considered in the future when more is known about targeting proteins to the chloroplast envelope.

1.4.TRANSFORMATION OF CHLAMYDOMONAB CHLOROPLASTS 1.4.1 Chlamvdomonas is a Unique Transformation System Genetic transformation of chloroplasts is the only way to evaluate irj. vivo the consequences of cpDNA alterations made in vitro and to explain and exploit changes in phenotype caused by presence of new sequences or the removal of pre-existing sequences in a chloroplast genome. Until recently, direct transformation of chloroplasts was difficult because this organelle is bounded by a double membrane envelope that may hinder transport and because chloroplasts are genetically polyploid. However, over the last three years a new transformation system (microprojectile bombardment) for the delivery of DNA capable of altering the phenotype of chloroplasts and mitochondria in living cells has been developed. The green alga Chlamvdomonas reinhardtii provides a unique system to study the transformation of chloroplasts by the application of the microprojectile bombardment. First, Chlamydomonas is a unicellular organism with short life cycle. These features make it easy to be handled in the laboratory similar to bacteria and yeast. Second, each Chlamvdomonas cell contains a single large cup-shaped chloroplast with a volume of 17-25 urn3 which occupies as much as 40% of the cell. Since the chloroplast lies adjacent to the plasma membrane along most of the periphery of the 10-um diameter cell, Chlamvdomonas reinhardtii is a favorable target for bombardment transformation by DNA-coated microprojectiles. Third, although the chloroplast in Chlamvdomonas is polyploid, with an average complement of about 80 genomes, it has a much lower genome copy number than other plant cells having many chloroplasts. Fourth, by growing cells in the presence of a DNA synthesis inhibitors the copy number of the chloroplast genome can be reduced without affecting the amount of nuclear DNA. For example, the chloroplast DNA copy number of Chlamvdomonas can be selectively reduced by the treatment with the thymidine analog, 5-fluorodeoxyuridine. This treatment leads to increased transformability (Boynton et al., 1988; Wurtz et al., 1977).

1.4.2 Delivery of DNA to Chloroplasts In 1987, a high velocity microprojectile bombardment device (particle gun) was first used to deliver macromolecules to plant cells by Klein et al. (1987). A generalized schematic of the gunpowder discharge apparatus described by them is shown in Fig 1. 21

gunpowder cartridge /'JTlrS

macroprojectiie 7* r micro be ads coated with DNA stopping plate

petri dish with cells

vacuum chamber

Figure 1. Schematic of a microprojectile delivery system adapted from Klein et al (1987). DNA coated tungsten particles are deposited on one face of a miroprojectile. The microprojectile is then accelerated against the stopping plate that has a , which allows the microbeads to bombard the cells in the vacuum chamber. 22 Using this device, Klein et al. (1987) have shown that small DNA-coated tungsten particles can be accelerated to velocities that permit their penetration of intact cells and tissues thereby introducing foreign DNA, which they subsequently showed was transiently expressed in intact epidermal cells of Allium ceoa (onion). Although activity through transient expression does not necessarily indicate stable transformants, this technology provides a simple, fast, and effective method for plant transformation. The most important feature of this technique is the ability of microprojectiles to pierce cell walls and cell membranes, which are the principal barriers to DNA delivery, without killing cells. Successful application of procedures (such as electroporation and cell fusion) to some plants has been limited by the inability to regenerate plants from transformed protoplasts and tissues. The microparticle technique, however, does not require regeneration from transformed protoplasts. Boynton et al. (1988) have reported the stable transformation to photoautotrophism of nonphotosynthetic, acetate-requiring of Chlamydomonas reinhardtii containing a 2.5 kbp deletion in atpB with wild-type cpDNA sequences delivered on microprojectiles. Southern blot analysis of photosynthetically active transformants demonstrated that the fragment with a 2.5-kilobase deletion was restored to normal size, presumably by homologous 23 recombination. Using the same nonphotosynthetic, acetate-requiring mutant of Chlamvdomonas reinhardtii. Blowers et al. (1989) further defined the chloroplast transformation system and used it for studying the role of chloroplast DNA sequences in gene expression. Their results showed that besides double-stranded circular plasmid DNA, single-stranded DNA circles, and linear duplex DNA molecules containing the wild-type atpB gene could also complement the deletion mutant via integration into the at the atpB . The transformation frequencies with linear DNA molecules were four to ten-fold greater than that observed for uncut circular DNA, suggesting that the free ends of the homologous chloroplast DNA stimulate recombination between the input DNA and the chromosomal DNA. Furthermore, a foreign, unselected chimeric gene flanked by chloroplast DNA sequences can be integrated and maintained stably in the Chlamvdomonas chloroplast genome. The bacterial neomycin phosphotransferase structural gene fused to the chloroplast for the large subunit gene of rbcL has been integrated into the inverted region of the Bam 10 restriction fragment. RNA transcripts that hybridize to the introduced foreign gene have been identified (Blowers et al. 1989). Recently, chloroplast transformations of cultured tobacco cells by the high-velocity microprojectile method have been reported, indicating that tobacco chloroplast transformation will likewise be achieved by bombardment 24 (Daniell et al., 1990; Svab et al., 1990). An alternative transformation system has also been developed recently. In this system, transformation can be accomplished by agitating cells In the presence of glass beads and DNA. This method was successfully used to transform Intact, cell wall containing yeast cells by Lostanzo and Fox (1988). However, the procedure for Chlamydomonas is not efficient unless the cell wall is removed, either by or by treating cell with autolysin. Kindle (1990) has reported that when cell-wall-deficient Chlamydomonas cells are agitated in the presence of glass beads, DNA, and polyethylene glycol (PEG), transformation of nuclear genome occurs at a high rate. Simply by agitating cell wall-deficient cells in the presence of glass beads and DNA, Kindle et al. (1991) have also recovered up to 50 chloroplast transformants per microgram of DNA in Chlamvdomonas. Compared to microprojectile bombardment system, the glass bead method has several advantages. It is simple, inexpensive and does not require specialized equipment. On the other hand, this method is not as efficient as the tungsten particle bombardment. This is probably because the shear force generated during agitation is either not sufficient to disrupt three membrane (plasma membrane, and the outer and inner chloroplast envelope membranes) to introduce foreign DNA into chloroplasts or may be too harsh causing the death of cells. It is also questionable whether this method will be useful for transformation of cells and organelles from 25 other organisms. The internal location and large numbers of chloroplasts in higher plant cells combined with larger cell size might render this method less useful for higher plant chloroplast transformation. Furthermore, the glass bead procedure requires cell wall-less cells. Although Chlamydomonas cells without walls are osmotically stable due to active contractile vacuoles, higher plant protoplasts would be unlikely to survive the shear forces generated in the glass bead procedure.

1.4.3 Selectable Markers for Chloroplast Transformation Since there is a large copy number of chloroplast genomes in the plant cells, experiments designed to establish stable chloroplast transformation require selectable marker genes encoded by the chloroplast genome. One kind of selectable marker is deletion or point mutations. If these mutations are known to eliminate chloroplast gene function, chloroplast transformants can be selected based on the ability of transforming DNA to complement the mutation. For example, photosynthetically defective mutants of the chloroplast atpB gene of Chlamydomonas reinhardtii were transformed with the cloned wild type chloroplast DNA sequence and the transformants were selected for their restored photosynthetic capacity. Southern blot analysis showed that in the transformants, a fragment with a 2.5-kilobase deletion was restored to normal size by a homologous replacement event 26 (Boynton et al., 1988). Antibiotic resistance markers inserted into the chloroplast genome could also be useful as tools for chloroplast transformation. Several selectable markers for chloroplast transformation that do not rely on selection for phototrophic growth have recently been developed by Boynton et al. (1990). A question is whether can be generally used to replace wild-type Ch1amvdomonas chloroplast genes with mutant ones. One strategy that has worked takes advantage of co-transformation with two different markers. It has been reported that cotransformation of two markers on independent plasmids into the nuclear genome is very efficient using either microprojectile bombardment (Kindle et al., 1989) or glass bead transformation (Kindle, 1990). In the chloroplast, markers conferring streptomycin resistance (encoded by 16s rRNA-encoding DNA) and erythromycin resistance (encoded by 23s rRNA-encoding DNA) are separated in the chloroplast genome by about 4.2 kilobases. When they were introduced into the chloroplast on single plasmid, about twenty percent of the streptomycin-resistant transformants were also erythromycin resistant (Boynton et al 1990). By using the spectinomycin resistance allele of 16s rRNA as the selected marker. Kindle et al. (1991) have taken advantage of co-transformation with in vitro-constructed mutant forms of the atpB gene. Among spectinomycin-resistant transformants, up 27 to 75% had also incorporated the introduced mutant atpB DNA. This result suggests that any chloroplast DNA segment altered in vitro can be introduced into the genome to make a mutant in a predictable way. Using a similar co-transformation system, a mutagenized psbA fragment has been introduced into the wild type Chiamvdomonas chloroplast genome, with subsequent conversion of the psbA-encoded photosystem II reaction center D1 protein (Roffey et al. 1991). It should also be noted that previous treatment with 5- fluorodeoxyuridine to reduce chloroplast DNA copy number is necessary for efficient chloroplast transformation and co­ transformation (Boynton et al., 1990; Kindle et al., 1991). This is likely to be especially important for the introduction of a recessive mutant which offers no advantage to the cell. In conclusion, the ability to genetically transform Chlamydomonas chloroplasts by introducing DNA into the chloroplast genome greatly enhances the utility of this organism for experimental research. The availability of selectable markers, combined with a workable transformation system provides a model for the study of chloroplast function, regulated gene expression and identification of unknown chloroplast genes (open reading frames). The model is especially valuable since efficient transformation of higher plant chloroplasts has not yet been achieved. 28 1,9 Rggft PR9TBIW 1.5.1 General Characteristics In 1965, Clark and Margulies isolated mutants of which are blocked in . A unique genetic locus responsible for these mutants is designated recA (Clark and Margulies, 1965; Clark, 1973). Further biochemical and genetic studies have shown that the of the recA gene is 1056 nucleotides long and encodes a protein of 352 amino acids with a molecular weight of 37842. The recA protein has alanine and as its NH2- and C00H- terminal amino acids, respectively (Horii et al., 1980). The predicted secondary structure and the approximate tertiary folding of recA protein suggest that it possesses a "nucleotide binding fold" and consists of six putative B-strands alternating with five putative a-helix regions (Blanar et al., 1984). However, the actual three- dimensional structure of recA protein is unknown at present. A striking feature of recA protein is its tendency to form aggregates or higher order polymers. Under appropriate conditions, recA protein self-assembles into a long filamentous structure which is visible in the electron (McEntee et al 1981, Flory and Radding 1982, Cotterill and Fersht 1983). The filament formation is very sensitive to ionic strength, and occurs in the absence of DNA. In addition, nucleotides (such as ATP and GTP) have been observed to disrupt recA protein filaments in both electron 29 microscopic and light-scattering studies (Cotterill and Fersht, 1983). These observations suggest that the DNA- independent formation of filaments is not on the normal reaction pathway for recA protein promoted DNA strand exchange. This is supported by the observation of competition between the formation of free recA protein filaments and the binding of recA protein to single-stranded DNA (ssDNA) (Morrical and Cox, 1985). This implies that the formation of functional recA protein complexes on ssDNA may involve the addition of a small unit rather than an intact filament. Therefore, free recA protein filaments do not form functional complexes with ssDNA and cannot be considered direct intermediates in recA protein-promoted DNA strand exchange reactions. As visualized by electron microscopy, recA protein binds to ssDNA in a highly cooperative manner at neutral pH (Dunn et al., 1982). In the presence of recA protein, added circular ssDNA was coated and extended significantly to form complexes of protein and DNA. The extended circular filaments are stiff and regular in appearance, contrasting with the convoluted structure formed by single strand DNA binding protein and ssDNA (Flory and Radding, 1982). Binding of recA protein to ssDNA is relatively stable as judged by its rate of equilibration with a challenging DNA (t1/2 of reassociation approximately 30 min) (Bryant et al., 1985). Addition of ATP to recA protein-ssDNA complexes 30 stimulates the equilibration of recA protein with added challenge DNA so that the t1/2 of reassociation is approximately 3 min. ADP causes an even greater stimulation of equilibration (t1/2 is about 0.2 min). These observations suggest that the slow equilibration leading to or including the ATP hydrolysis step precedes a rapid ADP-induced release from the DNA. Based on these studies, the role played by recA protein-ssDNA complexes in recA protein-promoted DNA-pairing reactions is very likely as an active intermediate and involves the following processes as suggested by Bryant et al. (1985) : recA initially interacts with ssDNA to form the recA- ssDNA complex. Binding of ATP to the complex then converts recA protein to a high-affinity conformation that can interact with a second strand of ssDNA. When ATP is hydrolyzed to ADP and Pi, recA protein is released from the ssDNA and the complex breaks down. With added challenge DNA, rebinding of recA protein to ssDNA occurs. Both single- and double-stranded DNA stimulate prehydrolytic recA binding to form recA-DNA complexes. However, binding of recA protein to duplex DNA is different from recA-ssDNA binding (McEntee et al., 1981; Weinstock et al., 1981). The two reaction also differ in their salt sensitivity and degree of inhibition by ADP. In the presence of ATP, recA protein unwinds the double helix of form I DNA to produce positive-superhelical turns which can be relaxed by eukaryotic topoisomerase I, resulting 31 in closed-circular dsDNA with an extraordinarily large number of negative-superhelical turns (Ohtani et al 1982). This ATP- dependent unwinding of the double helix might promote dissociation of D-loops in form I DNA, followed by inactivation of the form I DNA and elongation of heteroduplex joints. Thus, recA protein might effectively the DNA at or near the transition state for exchange of strands. ATPase activity was found to be associated with recA protein. Both ssDNA and dsDNA stimulate the hydrolysis of ATP catalyzed by the recA protein of £ coli (Ogawa et al., 1978? Roberts et al., 1978). However, the reactions differ in their pH optima; the dsDNA-dependent reaction exhibits a pH optimum near pH 6.0 while the ssDNA-dependent reaction shows a broad pH optimum between 6.0 and 9.0 (Weinstock et al., 1981). Since ssDNA and dsDNA stimulate hydrolysis of same nucleoside triphosphates, principally ATP and UTP, a single hydrolytic site could be utilized in both ssDNA and dsDNA-dependent reactions. Another striking property of recA protein is its ability to renature complementary single strands (Weinstock et al., 1979). The renaturation reaction catalyzed by recA protein is stimulated by ATP (Bryant and Lehman, 1985; McEatee, 1985). The two fold increase in the stoichiometry of binding of recA protein to ssDNA in the presence of ATP over that seen in the absence of ATP suggests that a recA monomer contains two DNA- binding sites (Weinstick et al 1985). The two DNA-binding 32 sites are also required for the mechanisms of transfer of recA protein from one ssDNA molecule to another. This cooperative transfer is stimulated by ATP and characterized by the intermediate complex between a recA-ssDNA and a second ssDNA molecule, followed by transfer of the recA protein from the first to the second strand (Bryant et al., 1985; Menetski and Kowalczykowski, 1987). Certainly, the formation of such a two- stranded intermediate is considered an important step to bring complementary DNA strands together so that pairing and strand exchange can occur.

1.5.2 Role in General Recombination Genetic recombination involves reciprocal exchange of homologous regions of DNA between two different DNA molecules resulting in recombinant DNA molecules containing genetic information originally present in each of parental molecules. In £. coli, genetic and biochemical studies demonstrated that recombination is a controlled process, affected by special genes and by the specificities of protein products promoting these process. One of the proteins, recA, has been shown to be essential to promote homologous pairing and strand exchange between DNA molecules in an ATP-dependent reaction (Clark, 1973; Cunningham et al., 1980; McEntee et al., 1979; Smith, 1987). The activities of recA protein described above, including binding to DNA, ATP hydrolysis, and renaturation of complementary DNA strands, converge in the DNA strand exchange 33 reaction. The substrate DNA strands that can be utilized by recA protein for strand exchange must have three features (Smith, 1987). One of them is nucleotide sequence homology between the substrates. The minimum required length of sequence homology is between 30 and 151 nucleotides. Another requirement is that at least part of one of the substrates must be ssDNA, and the ssDNA region must occur at a site homologous to the other dsDNA molecule. For topological reasons, a DNA end is also necessary to be in one of substrates. The reaction can then take a variety of forms and lead to a variety of products (Figure 2). The simplest reaction is the annealing of complementary single strands (Figure 2a). More complex reactions will involve duplex DNA if one of the two molecule is single-stranded (Figure 2b) or partially single-stranded, either with single-stranded gaps (Figure 2c) or single­ stranded tails (Figure 2d). The sequence of events leading to DNA strand exchange include: binding to DNA substrates, pairing of the DNA substrates, search for homologous regions, local disruption of dsDNA structure and DNA strand exchange. Thus, the whole reaction can be divided into at least three phases. In the first presynaptic phase, the recA protein polymerizes on ssDNA to form a nucleoprotein filament. Direct ssDNA binding studies have demonstrated that presynaptic filaments, when mixed with duplex DNA and ATP, form joint 34

(a) (b) (c) (d)

Figure 2. Reactions promoted by recA protein (Smith 1987). ssDNA and dsDNA are represented by single and double lines, (a) Annealing of complementary ssDNA. (b) D-loop formation, (c) A widely studied strand-transfer reaction, (d) Reciprocal strand-transfer and . 35 molecules more rapidly and efficiently than equivalent concentrations of free recA protein and ssDNA (Flory et al., 1984; Tsang et al., 1985). As mentioned above, ATP does have a striking effect on the equilibrium binding affinity of recA protein to the DNA. Both ATP and the non-hydrolyzable ATP analogue, ATPrs, significantly increase the affinity of recA protein for ssDNA (Menetski and Kowalczykowski, 1985). In contrast, the effect of ADP is to decrease the stability of the recA protein-ssDNA complex. These results suggest that the ATP hydrolytic cycle modulates the affinity of recA protein between two different ssDNA affinity states and thereby facilitates the cyclic binding and dissociation of recA protein from ssDNA (Menetski and Kowalczykowski, 1985). The binding of recA protein to ssDNA is polar, proceeding in a 5* —3' direction (Register and Griffith, 1985). This is the same direction as strand assimilation during strand exchange. Although at the simplest level presynapsis involves the binding of recA protein to ssDNA, the filamentous protein- nucleic acid complexes are actively involved in the subsequent phases of strand exchange. The second phase of the reaction involves alignment of the ssDNA within the nucleoprotein filament with complementary sequences in the duplex DNA. As mentioned above, the binding of ssDNA stimulates recA protein to bind duplex DNA. Whether the two forms of DNA are homologous or not, the synapsis phase 36 appears to consist of two sequential steps: the conjunction of DNA molecules and their homologous alignment (Gonda and Radding, 1983). A number of studies have provided evidence that in the presence of ATP, recA protein-ssDNA complexes can bind nonspecifically to heterologous duplex DNA (Shibata et al., 1979; Tsang et al., 1985; Chow and Radding, 1985). These nonspecific interactions are manifested in vitro by large nucleoprotein networks that link together many recA protein- ssDNA presynaptic filaments and duplex DNA molecules early in the strand exchange reaction (Chow and Radding, 1985). The nature of these complexes and their role in homologous pairing in vivo is not very clear. However, when naked duplex DNA is added to fully formed presynaptic filaments under conditions that are suitable for homologous pairing, both ssDNA and dsDNA rapidly coaggregate (Tsang et al., 1985; Rusche et al, 1985). It has also been demonstrated that the pairing of short duplex molecules with single strands is accelerated by the addition of long heterologous duplex molecules (Gonda and Radding, 198 6), suggesting that longer duplex molecules make more stable coaggregates by making more contacts. These results are consistent with the concept of two binding sites per molecule of recA protein. Thus, the presynaptic filament can form large nucleoprotein networks by nonspecifically binding duplex DNA at many "second sites" as it searches for homology. Since these nucleoprotein networks bring DNA molecules into proximity, they provide intermediates that speed homologous 37 alignment by concentrating the DNA and lead to strand exchange reactions. During the homologous alignment a minimum required length of sequence homology may be necessary to discriminate between heterologous and homologous contacts. Based on the rate of formation of stable joint molecules by circular single strands and linear duplex DNA molecules, Gonda and Radding (1983) found that the minimum required length of sequence homology is between 30 and 151 nucleotides. Another interesting study about the effects of mismatches within a homologous region was done by Watt et al (1985), who observed a 10-fold decrease in recombination resulting from a single mismatch within a homologous region of 53 base pairs. After homologous alignment completes the synapsis, the strand exchange phase occurs. The single strand originally coated by recA protein replaces its homologous strand in the duplex DNA to form a new heteroduplex region or molecule. After the single-stranded and duplex DNA molecules are homologously aligned, the mechanism of local denaturation of the duplex DNA molecule and subsequent exchange of the identical ssDNA must be considered. Since the local denaturation of dsDNA requires the input of energy, it is reasonable to suspect the involvement of ATP in this process. In fact, recA protein can processively polymerize onto dsDNA and induce a topological unwinding of the dsDNA in the presence of ATP (Shibata et al., 1984). This observation is also consistent with the dsDNA-dependent ATPase activity of 38 recA protein (Kowalczykowski et al., 1987). Continued ATP hydrolysis is also required throughout the branch migration process, although there is no correlation between the rate of ATP hydrolysis and the number of migrating branch points (Brenner et al., 1987). Clearly, more extensive studies are required to understand the role of ATP binding and hydrolysis in strand exchange. Strand exchange proceeds with a preferred chemical polarity. The reaction of linear ssDNA and circular dsDNA requires a 3 ■ end on the ssDNA homologous to the dsDNA (Konforti and Davis, 1987). The reaction of linear dsDNA with circular ssDNA requires a 3' end on the dsDNA that is homologous and complementary to the ssDNA (Cox and Lehman, 1981). The polarity of reaction has been defined as initiating at a 3*end if the substrates are considered (Kondorti and Davis, 1987). The reaction scheme of DNA recombination catalyzed by £. coli recA protein presented in Figure 3 is based on the one described by Cox and Lehman (1987). Synapsis including nonspecific binding between ssDNA and dsDNA followed by homologous alignment is shown as the first step (a). In the strand exchange phase, nonhomologous interactions lead to formation of a paranemic joint in which the individual complementary strands do not intertwine, resulting in a molecule that is base paired though not topologically linked (b) . In the third step (c) heteroduplex joints are induced by ATP

♦

Figure 3. Reaction pathway for the transfer of a strand from a linear duplex to a circular ssDNA (Cox and Lehman, 1987). It shows (a) nonhomologous interactions leading to (b) formation of a paranemic joint. Formation of a plectonemic joint (c) is coupled to ATP hydrolysis. Branch migration (d) is shown as a step separate from plectonemic joint formation, but is most likely an extension of the same process. 40 the ATP-recA in the plectonemic form in which the incoming single strand of DNA is intertwined around its complements as in native dsDNA, followed by strand migration (d) . As such, recA plays the central role in homologous recombination as a .

1.5.3 Role in DNA Repair The recA gene of £. coli plays an essential role in a variety of processes that enhance cellular survival following exposure to DNA-damaging agents (Witkin, 1976). An indication of the critical importance of recA gene product in £. coli cell physiology is the pleiotropic nature of recA mutations. Study of £. coli recA mutants has revealed that such cells are significantly more sensitive to UV radiation and other DNA damage agents than their isogeneic parents (Clark, 197 3). For example, UV radiation can create certain nucleotide adducts, such as dimers, which blocks replication about 1000 bp beyond the adducts, thus generating a single-stranded gap. This discontinuity can be filled in by the recA protein associated repair system in normal cells. In uvr' recA' cells, attempts to replicate DNA produce DNA fragments of a size corresponding to the expected distance between thymine dimmers (Lewin, 1990). This implies that the dimer provides a lethal obstacle to replication in the absence of recA function. The following model describes a likely sequence of events in recombinational repair in £. coli (Howard-Handers et al., 41 1984). The recA protein binds to duplex DNA carrying a single­ stranded gap and forms a long helical complex which pairs with intact homologous duplex. Reciprocal strand exchange will then take place, partly facilitated by polymerization from the free 3' terminus of the damaged duplex. The final products will be the two uninterrupted duplexes. In addition to its direct participation in DNA repair by recombination, the recA protein is required for induction of the expression of other repair genes subsequent to DNA damage or inhibition of DNA replication in £. coli (Kenyon and Walker, 1980; Kenyon et al., 1982). The responses induced by DNA damage include: mutagenesis, inhibition of cell division, and increased DNA repair capacity. This complex series of phenotypic changes has been termed the "SOS response" (Radman, 1975) . The response takes the form of an increased capacity to repair damaged DNA, achieved by the SOS regulatory system involving the recA and lexA gene products (Little and Mount, 1982). As shown in Figure 4, the lexA protein serves as the direct repressor of all SOS genes by binding to sequences of ten bases referred to as SOS boxes, that are located near their promoters (Brent and Ptashne, 1981; Little and Mount, 1982). In untreated normal cells, most SOS genes are expressed at a basal level even in the absence of an SOS-inducing treatment. The uninduced level of expression of an SOS gene may be influenced by the physical relationship of the promoter to the lexA-binding sequence (Backendorf et al., 198 3) or the 42

^ uvrA

/-n r*cN

^ umuOC FD

uvrA J------U

rtcN ■l. , j_ rtcA

utnuDC J------L

SOS Inducing signal

Figure 4. Model of the SOS regulatory system (Walker, 1985). LexA is the repressor of SOS genes. The generation of an inducing signal as a consequence of DNA damage leads to activation of recA. The interaction of activated recA with lexA results in the cleavage of lexA. As the lexA pools decrease the SOS genes (such as uvrA, recN, umuDC) are expressed at higher levels. 43 presence of an additional unregulated promoter (Sancar et al., 1982). The initial event in the SOS response is the activation of recA by the damage treatment. Although not much is known about the relationship between damage events and the sudden change in recA activity, a common inducing signal could exist. On activation, the recA protein interacts with the repressor protein coded by the lexA gene. This interaction with activated recA triggers proteolytic cleavage of the repressor at a specific alanine-glycine peptide bond near the midpoint of the lexA polypeptide structure (Little et al., 1980). The decrease in the pool of intact lexA protein increases expression of all SOS genes, including the recA itself, which allows more efficient DNA repair (Peterson and Mount, 1987). Studies of purified lexA protein from £. coli have shown the presence of an autoproteolytic activity under certain conditions, although the site of cleavage is the same as in the recA-mediated reaction (Little, 1984; Slitaty et al., 1986) . This implies that recA may not directly cleave the peptide bond of lexA. Rather, interaction with activated recA may change the conformation of lexA protein in a manner which facilitates its autodigestion (Walker, 1985). Although the exact mechanism of this cleavage is unknown, the recA protein must be activated as a coprotease to accelerate the autocatalytic process. 44 1.5.4 RecA Protein Analogues Proteins with properties similar to recA protein of £. coli appear to be widely distributed among bacteria, including Gram-Negative bacteria (Eitner et al., 1982; Keener et al., 1984) and Gram-positive bacteria (Lovett and Roberts, 1985). In photosynthetic , recA analogues have been identified in Gloecapsa alpicola (Geoghegan and Houghton, 1987), Svnechococcus sp (Murphy et al., 1987) and Anabaena variabilis (Owttrim and Coleman, 1987). The recA clones from all three genera of cyanobacteria can complement defects in homologous recombination and restore UV resistance to £. coli recA mutants (Geoghegan and Houghton, 1987; Murphy et al., 1987; Owttrim and Coleman, 1987). A recA-like protein (reel) has also been purified from mitotic cells of the lower Ustilaao mavdis based on its ability to reanneal complementary single strands of DNA (Kmiec and Holloman, 1982). This protein also promotes the uptake of ssDNA by duplex DNA in a homology dependent reaction. ATP is required and hydrolyzed in this reaction. However, the polarity of branch migration catalyzed by reel is opposite of recA protein (Kmiec and Holloman, 1984). Several groups have attempted to detect recA-like activities in mammalian cell extracts. Kenne and Lindquist (1984) have reported the formation of joint molecules of superhelical DNA and single-strand fragments in fibroblast extracts. A strand transfer activity also has been partially 45 purified from human B lymphoblast (Hsieh et al., 1986). Unlike E- coli recA protein, the reaction catalyzed by the lymphoblast protein does not require ATP. In contrast, the strand transfer and homologous pairing reactions are dependent on the presence of ATP and requires homology between DNA substrates in human fibroblasts and Hela cells (Cassuto et al., 1987). In extracts of FR 3T3 cells, a protein which immunologically cross-reacts with the E* coli recA protein has been found (Angulo et al., 1989). There is no report of a recA-like protein in plants. However, a chloroplast DNA sequence which is homologous to the £. coli recA gene has been identified in the Chlamvdomonas chloroplast genome (Oppermann et al., 1989), suggesting that chloroplast DNA possibly encodes a recA-like protein. The presence of such a protein in the chloroplast has obvious implications for the introduction and integration of foreign DNA into chloroplasts. CHAPTER II

REDUCTION 07 CHLOROPLAST DNA CONTENT IN SOLANUM NIGRUM

SUSPENSION CELLS BY TREATMENT WITH CHLOROPLAST DNA

SYNTHESIS INHIBITORS

2.1 INTRODUCTION Previous investigators have reported that cpDNA synthesis can be selectively inhibited by a variety of compounds. In Eualena. inhibition of cpDNA synthesis by nalidixic acid, a prokaryotic DNA gyrase inhibitor, has been shown to lower the steady state copy number of chloroplast genomes without affecting the nuclear DNA content (Lyman, 1967; Lyman et al., 1974; Pienkos et al., 1974). In the green alga Chlamvdomonas. treatment with the thymidine synthesis inhibitor Fudr and/or the DNA binding dye ethidium bromide results in the specific reduction of chloroplast DNA content which recovers to normal levels following removal of the inhibitor (Flechtner and Sager, 1973; Wurtz et al., 1979). In addition, Fudr treatment of Ch1amvdomonas has also been shown to increase the level of transmission of chloroplast DNA for mt' parent in crosses and recovery of chloroplast mutations (Wurtz et al., 1979). This feature, reduction in chloroplast genome number, has been exploited as a means to increase the frequency of chloroplast

46 47 transformation in Chlamvdoroonas (Boynton et al., 1990). Unfortunately, there have been few studies on the manipulation of chloroplast DNA content in higher plants. In one of the few studies involving higher plants, Weisbach and coworkers (Heinhorst et al., 1985, 1986) demonstrated that nalidixic acid inhibited chloroplast DNA synthesis in higher plant suspension cultures. However, it was not determined whether nalidixic acid treatment was toxic or, in fact, lowered steady-state levels of cpDNA. To determine whether cpDNA steady-state levels could be effectively reduced in higher plants without causing cell death, we screened a number of potential cpDNA synthesis inhibitors in suspension cell cultures of Solanum nigrum for their effects on chloroplast and nuclear DNA content and cytotoxicity. These studies were carried out with cultures which were either grown to stationary phase or rapidly transferred so as to maintain them in an active state of cell division and cpDNA replication (Yasuda et al., 1988). The results of our experiments demonstrate that certain inhibitors can selectively lower the steady state level of chloroplast DNA in actively dividing cultures with minimal effects on culture viability. It is proposed that such treatments may be exploited as means to facilitate the generation and segregation of cpDNA mutants and/or cpDNA transformants in higher plants. 48

2.2 MATERIAL8 AND METHODS 2*2.1 Plant Material and Growth Conditions Seeds of wild type Solanum nigrum L. were surface- sterilized in 10% Clorox for 20 min, rinsed three times with sterile water, and germinated on MS medium (Hurashige and Skoog, 1962) at 28°C. Excised hypocotyl sections from 10 to 14 day old seedings were used to initiate fresh callus cultures by culturing on MS medium supplemented with 2 mg/L 2,4-D and 0.2 5 mg/L kinetin at monthly intervals. Suspension cell cultures were initiated by transfer of 2 g callus to 50 ml of liquid MS (supplemented) medium and maintained on a shaker in the dark at 28°C. Suspension cultures were subcultured by transferring 10 ml of suspension to 50 ml of fresh medium at 5 day intervals.

2.2.2 Isolation of Protoplasts Protoplasts were isolated by digestion of 4 g of suspension cells in 15 ml of 2% (w/v) cellulase (Onozuka RS), 0.2% (w/v) macerozyme (Onozuka R-10), 0.5 M mannitol, and 0.1% CaCl2-2H20. Cells were incubated on a platform shaker at 30 rpm for 14 to 16 h. The protoplasts were isolated by filtration through 200 um and 74 urn sieves, washed two times with 0.5 m mannitol and harvested by floatation on a 20% sucrose solution which was centrifuged at 100 x g for 10 min. Protoplasts were then collected and rinsed twice with 0.5 M mannitol. 49 2.2.3 Treatment of Suspension Cells with Chloroplast DNA Synthesis Inhibitors Suspension cells were broken up into small clumps by vortexing in an impingement tube and then filtered through a 20 mesh (890 um) sterile stainless steel wire cloth. The homogenous cells were incubated in fresh MS (supplemented) medium for 2 days to allow the cells to reach early log phase. One gram (fresh weight) of cells was then transferred to 50 ml fresh MS (supplemented) medium plus 1 g casein and one of the following potential chloroplast DNA synthesis inhibitors: bisbenzymide, ethidium bromide, Fudr, nalidixic acid, novobiocin, and . The cell cultures were incubated with inhibitors for 8 days at 27 to 28°C and (fresh weight) was determined. The cells were then pelleted, frozen in liquid nitrogen, and stored at -80°C prior to DNA extraction. In another set of culture conditions, 2 g of inoculum were grown in fresh medium plus inhibitor for 4 days (it was necessary to use 2 g of tissue instead of 1 g due to cell death at higher concentrations of the inhibitors) and then transferred (1 g) to new culture medium plus inhibitor for 4 more days growth prior to DNA extraction. The effect of inhibitors on cell viability was determined by the recovery rate of cell growth after transfer to inhibitor free medium. Treated cells were washed and then grown in fresh medium for 7 days prior to fresh weight determination. Rates of growth were compared to cultures which had not been treated with 50 inhibitors.

2.2.4 Analyses of Chloroplast and Nuclear DNA Content Three different DNA isolation procedures (Dellaporta et al., 1983; Laulhere and Rozier, 1976; Murray and Thompson, 1980) were compared for yield of chloroplast and nuclear DNA and degradation of DNA. The procedure that proved to be most effective for complete extraction of intact nuclear and chloroplast DNA was a modification of the mini-prep procedure developed by Dellaporta et al. (1983). Briefly, l g of tissue was ground in liquid nitrogen to a fine powder. The dry powder was placed in 15 ml of extraction buffer containing 0.1 M Tris-HCl (pH 8.0), 0.05 M EDTA, and 0.01 M B-mercaptoethanol. After adding 1.0 ml of 20% SDS, the solution was mixed thoroughly by vortexing and heated at 65°C in a water bath for 10 min. Five ml of 5 M potassium acetate was added, and the solution was incubated on ice for 20 min to precipitate proteins and polysaccharides. These contaminants were removed by pelleting at 2,500 xg for 20 min. The supernatant was then poured through sterile miracloth into a 30 ml vertex tube containing 10 ml of isopropanol and placed at -20°C for 30 min. Total nucleic acids were pelleted at 20,000g for 15 min, resuspended in 600 ul water, and treated with 1.0 mg of RNase A for 1 h at 37°C. After phenol and CHC13 extraction, the DNA was precipitated with 1 volume of isopropanol and 0.1 volume of 3 M sodium acetate at -20°C overnight. The precipitate was 51 then pelleted In a microfuge for 10 min, washed with 70% ethanol, and resuspended in 300 ul TE buffer (10 mM Tris-HCl, pH 8.0, and 1 mM EDTA). Total extracted nucleic acids were bound to nitrocellulose using the procedure recommended by Schleicher and Schuell. DNA was loaded on the basis of tissue fresh weight extracted or cell number. The DNA (100 ul) was denatured by adding 15 ul of 3.0 M NaOH and 35 ul of water followed by incubation at 68°C for 1 h. After cooling to room temperature, 1.0 volume of 2 M ammonium acetate was added and mixed. One-half of each sample was applied under vacuum to a nitrocellulose membrane (0.45 urn) which had been prewashed in 1 M ammonium acetate. Known amounts of chloroplast and nuclear DNA standards corresponding to 30, 100, 300 and 1,000 ng of plasmid pBM5 (Mulligan et al., 1984) containing the psbA gene (chloroplast marker) and plasmid pSSU-160 containing the nuclear encoded small subunit of ribulose-1,5 bisphosphate carboxylase (Bedbrook et al., 1980) were immobilized on the same membrane in order to confirm the linearity of dot blot signals. The membrane was then baked for 2 h at 80°C in a vacuum oven. Membranes were prewashed in 200 ml of 50 mM Tris-HCl (pH 8.0), 1 M NaCl, 1 mM EDTA, and 0.1% SDS for 1 h at room temperature prior to prehybridization in a heat seal bag containing 4 ml of 5 X SSPE (0.15 M NaCl, 0.01 M NaH2P04, 0.001 M EDTA pH 7.4), 0.1% SDS, 5% dry milk, and 100 ug denatured 52 calf thymus DNA at 40°C overnight. DNA fragments encoding the psbA and pSSU-160 genes were labeled by the random primer extension method (Feinberg and Vogelstein., 1983). Approximately 3 X 106 cpm of denatured probe was then added to the bag and incubated for 20 h at 40°C. Filters were then washed three times in 2X SSC containing 0.1% SDS at room temperature, followed by three washes with IX SSC containing 0.1% SDS at 65°C for 1 h. The filters were then autoradiographed and signals quantified by densitometry.

2.2.5 Quantification of Total Nucleic Acids bv Fluorescent DNA Binding Dves Equal numbers of protoplasts (3 X 10s cells/ml) from each treatment were stained with 1 ug/ml DAPI (4,6-diamido-2- phenylindole) dissolved in Suzuki-Nishibayashi buffer as described in Miyamura et al. (1986). Examination of labeled protoplasts using a fluorescence microscope indicated that virtually all fluorescence was localized in either the nucleus or chloroplasts. Relative nucleic acid content of DAPI-labeled protoplasts (3 X 105 cells/ml) was determined spectrofluorometrically using an excitation wavelength of 372 nm and an emission wavelength of 456 nm and expressed on the basis of fluorescence values from untreated cells. 53

2,3 RESULTS 2.3.1 Effects of DNA Synthesis Inhibitors on Cell Growth and viability The chloroplast DNA synthesis inhibitors tested included: nalidixic acid and novobiocin, chloroplast DNA gyrase inhibitors (Gellert et al., 1976,1977; Heinhorst et al., 1985; Lam and Chua, 1987; Lockshon and Morris, 1983; Sugino et al., 1977,1978), rifampicin, an inhibitor of chloroplast RNA (Surzycki, 1969; Wehrli et al., 1968), ethidium bromide and bisbenzimide, DNA template binding dyes that inhibit chloroplast DNA synthesis (Flechtner and Sager, 1973; Heinhorst et al., 1985, 1986), and 5-fluorodeoxyuridine, which inhibits thymidine synthesis in Chlamvdomonas chloroplasts Wurtz et al., 1979). Cultures were either grown to stationary phase in continuous culture (8 d) or maintained in log phase of growth by transfer to fresh media on day 4 of the 8 day cycle (Fig. 5) . The selection of inhibitor concentrations used for these studies was based on their effects on culture growth and viability. The effects of various inhibitor concentrations on culture growth are shown in Figure 6. Treatment with 100 uM bisbenzimide, 8 uM ethidium bromide, 0.03 uM fluorodeoxyuridine, 200 uM nalidixic acid, 30 uM novobiocin, and /or 10 ug/rol of rifampicin inhibited cell growth by 50% in cultures maintained for 8 d in the same culture medium. Interestingly, several inhibitors were apparently more toxic 54

8 Days Culture 2 x 4 Days Culture

to - E E o* O’

JZ o*

Culture Days Culture Days

Figure 5* Growth of £. nigrum cultures in culture medium for 8 continuous days or following transfer on day 4 to new medium fcfr 4 more days (2X4 day) ; see "Materials and Methods'* 55

too too

(J§ |» *0 * £

b .

K SO »00 3 0 0 fiitbtniitnid* (j*M) Nolidiiic Add (fJU)

CO too

S *o

4 0 t£

Elbr [kM1

too too £ •o uI I* I 40 •0 ! »•

OOt 0 03 O i 0 3 LO 10 3 0 Fudr Ail

Figure 6. Effects of inhibitors on growth of cell suspensions of S. niorum. (o), 8 d; (□), 2 x 4 d; (•), recovery after transfer to inhibitor free media (8 d) ; (■), recovery after transfer to inhibitor free media (2 x 4 d). 56 in rapidly dividing cultures (transferred at 4 day intervals to new medium plus inhibitor, see Fig. 5 growth rates) than in cultures that were maintained for 8 days in the same medium. Those inhibitors which were more toxic to cultures which were transferred once included bisbenzimide, ethidium bromide, and 5-fluorodeoxyuridine (Fig. 6) . Transfer of novobiocin treated cells to new medium during the 8 day treatment had no effect on culture growth. To determine whether exposure to the chloroplast DNA synthesis inhibitors was lethal to the cultures, an inoculum of inhibitor treated cells was transferred to inhibitor free medium to determine whether growth would resume at normal rates. In general, treatments which substantially reduced growth rates, i.e. 5-fluorodeoxyuridine, ethidium bromide, bisbenzimide, and rifampicin also suppressed growth upon transfer to inhibitor free medium (Fig. 6) . However, the toxicity of some inhibitors was greater than others. Ethidium bromide was the most toxic of the compounds tested followed by 5-fluorodeoxyuridine, bisbenzimide and rifampicin. In contrast, concentrations of novobiocin and nalidixic acid that reduced culture growth by 50% did not significantly affect culture growth following transfer to new media without inhibitor (Fig. 6). 57 2.3.2 Effects of DNA Synthesis Inhibitors on Nuclear and Chloroplast DNA Content The nuclear and chloroplast DNA content per g fresh weight of treated cultures was determined by dot blot hybridization using chloroplast (psbA) and nuclear (pSSU) gene specific probes. Chloroplast and nuclear DNA was analyzed from cultures grown at inhibitor concentrations which effectively reduced growth by approximately 10, 50, and 80%. The results from these measurements from 8 day and 2 X 4 day treatment are shown in Tables 1 and 2, respectively. With the exceptions of bisbenzimide and novobiocin, the chloroplast and/or nuclear DNA content of 8 day treated cultures was reduced 50% or more by each of the inhibitors tested (Table 1) . However, the concentrations of inhibitors which reduced the content of nucleic acids by 50% or more also reduced culture growth. Interestingly, the effectiveness the inhibitors varied with different culture conditions. The two DNA gyrase inhibitors, nalidixic acid and novobiocin, were more effective in reducing chloroplast DNA content in cultures which were transferred during the 8 day growth period, whereas rifampicin and ethidium bromide preferentially reduced chloroplast DNA content in 8 day continuous cultures and not in transferred cultures. In contrast, 5-fluorodeoxyuridine reduced chloroplast DNA content in both 8 day continuous and 2 X 4 day cultures. Only one of the inhibitors tested, bisbenzimide, did not reduce chloroplast DNA content under any conditions Table 1. Effects of DNA synthesis inhibitors on the DNA content/g fresh weight of £. nigrum suspension cells (8 day culture). Cells were cultured for 8 days with various concentrations of inhibitor. DNA extracted from these cells was applied to nitrocellulose and probed with chloroplast (psbA) and nuclear (ssRubisco) probes. Abundance of chloroplast and nuclear DNA was determined by densitometry of autoradiographies. The experiments were done in triplicate.

Inhibitor Concentration Nuclear DNA cpDNA

uM % of Control

None 100 + 3 . 3 100 + 3.6 Bisbenz imide 10 67 + 4 . 3 105 + 5.3 30 69 + 3.2 63 + 4.9 100 71 + 3.7 83 + 6.6 Ethidium bromide 2 51 + 3.9 33 + 3.4 4 61 4.4 45 + 3.6 8 49 + 3.3 30 + 3.5 Fluorodeoxyuridine 0.1 78 + 4.9 90 + 3.1 0.3 57 + 4.6 28 ± 2.9 1.0 57 ± 3.5 35 ± 4.3 Nalidixic acid 50 84 + 6.2 100 ± 4.1 100 49 + 5.0 53 ± 5.7 200 53 + 2.9 63 ± 6.1 Novobiocin 20 84 + 4.9 80 ± 5.6 30 82 + 3.8 80 + 4.1 40 73 ± 5.0 73 + 2.4 Rifampicin 3 ug/ml 69 + 4.8 55 + 4.0 10 61 + 4.3 30 ± 4.2 30 45 ± 4.9 25 + 2.8 Table 2. Effects of DNA synthesis inhibitors on the DNA content/g fresh weight of £. nigrum suspension cells (2X4 day cultures). Cells were cultured for 2 X 4 day (one transfer to fresh medium) with various concentrations of an inhibitor DNA extracted from these cells was applied to nitrocellulose and probed with chloroplast (psbA) and nuclear (ssRubisco) probes. Abundance of chloroplast and nuclear DNA was determined by densitometry of the autoradiographies. The experiments were done in triplicate.

Inhibitor Concentration Nuclear DNA cpDNA

uM % of Control None 100 + 4.1 100 + 3.7 Bisbenzimide 10 81 + 3.5 81 + 5.7 30 76 + 5.6 81 + 4.3 100 64 + 5.2 61 + 3.7 Ethidium bromide 2 81 + 5.1 106 + 4.7 4 79 + 4.5 84 + 4.3 8 69 + 4.7 65 + 6.2 Fluorodeoxyuridine 0.1 93 + 5.3 103 + 3.7 0.3 50 + 5.2 39 ± 5.1 1.0 40 + 4.1 26 ± 5.2 Nalidixic acid 50 55 + 5.4 42 ± 4.3 100 38 + 3.8 32 ± 3.9 200 36 + 2.2 19 ± 4.3 Novobiocin 20 98 + 6.1 94 ± 3.3 30 67 + 3.3 65 ± 4.5 40 71 + 5.2 48 ± 3.0 Rifampicin 3 ug/ml 74 + 4.0 65 ± 3.7 10 43 + 5.4 45 ± 4.3 30 48 + 3.9 35 ± 2.1 60 tested. The differences in effectiveness of inhibitors in reducing chloroplast DNA content in 8 day versus 2 X 4 day cultures may be due to selective inhibition of chloroplast DNA replication following transfer of cultures to new media (Yasuda et al., 1988) . As shown by Yasuda et al. (1988) , there is a burst in chloroplast DNA synthesis during the first 2 day after transfer of suspension cells to new media, whereas nuclear DNA synthesis rates do not peak until several days later. At inhibitor concentrations which reduced culture growth by 80% or more, each inhibitor tested substantially reduced the content of nuclear DNA (ranging from 30-65%) as determined by dot blot hybridizations expressed on the basis of fresh weight (Tables 1 and 2). The loss of nuclear DNA can not be readily accounted for by a reduction in the copy number of genomes similar to the reduction in chloroplast DNA content. However, as demonstrated in Figure 2, high concentrations of inhibitors can be lethal and may result in the generation of heterogeneous populations of viable and nonviable cells. Since it was not apparent whether determination of DNA content on the basis of fresh weight was biased due to the presence of dead cells, alternative methods of DNA quantification were used to determine independently the total DNA content of viable cells only. Protoplasts were prepared from 2 X 4 day inhibitor treated suspension cells and the relative amounts of total nucleic acids from the different treatments was 61 determined spectrofluorometrically using the fluorescent DNA binding dye, DAPI. As shown in Table 3, bisbenzimide, 5- fluorodeoxyuridine, and rifampicin treatments reduced the total nucleic acid content of protoplasts by less than 10% although they were shown to reduce chloroplast and nuclear DNA content/g fresh weight by 25 to 60% as determined by dot blot hybridizations. In contrast, ethidium bromide, nalidixic acid, and novobiocin reduced total nucleic acid content per cell by 20% or more as determined by DAPI fluorescence values (protoplasts). Since DAPI measurements do not discriminate between chloroplast DNA and nuclear DNA content, nuclear and chloroplast DNA per unit cell (protoplast) was determined by dot blot analyses using DNA extracted from protoplasts. As shown in and Figure 7 and Table 4, and the nuclear DNA content per cell of inhibitor-treated cultures was nearly identical to that of untreated cultures on a per cell basis. These results were in contrast to results expressed on the basis of fresh weight (Tables 2 and 3) indicating that measurements of DNA content on the basis of fresh weight were biased probably due to the presence of nonviable cells. Chloroplast DNA content expressed on the basis of cell number was not affected (less or equal to 15%) by 5-f luorodeoxyuridine, ethidium bromide, or rifampicin treatment. However, the two DNA gyrase inhibitors, nalidixic acid and novobiocin, reduced chloroplast DNA/cell by 35 and 45%, respectively. Significantly, novobiocin and 62 Table 3. Quantification of Total DNA in Protoplasts Labeled with the Fluorescent DNA Binding Dye, DAPI. Cultures were transferred once during the 8 day treatment (2X4 day) . Inhibitor concentrations were: 30 uM bisbenzimide, 4 uM ethidium bromide, 0.3 uM 5-fluorodeoxyuridine, 100 uM nalidixic acid, 30 uM novobiocin, and 10 ug/ml rifampicin.

DAPI Fluorescence Values Treatment Control Exp. 1 Exp. 2 Average

% Control 121 130 126 100 Bisbenzimide 127 135 131 104 Ethidium bromide 97 101 99 79 Fudr 115 119 117 93 Nalidixic acid 96 99 98 78 Novobiocin 102 104 103 82 Rifampicin 111 116 113 90 Figure 7. Dot. blot of a dilution series of DNA extracted from S. niarum protoplasts cultured for 2 X 4 days with various inhibitors and probed with the chloroplast encoded psbA gene (lane 1 and lane 3) or the nuclear encoded pSSU gene (Lane 2 and lane 4). The treatments are: A, control; B, 30 uM bisbenzimide; C, 4.0 uM ethidium bromide; D, 0.3 uM 5- fluorodeoxyuridine; E, control; F, 100 uM nalidixic acid; G, 40 uM novobiocin; H, 10 ug/ml rifampicin. Approximately 18.0, 10.8, 7.2, and 3.6 ug of total DNA were blotted in a dilution series labeled 5X, 3X, 2X, and IX, respectively.

63 64

1

E. 5X

3X

2X

1X

■■ B. F. 5X 5X

3X — — 3X w * 2X ------2X I ttm.

1X 1X

C. G. 5X 5X

3X — — 3X

2X — ------2X

- 1X ------1X

D. H. 5X SX

3X 3X

2X 2X

1X 1X

Figure 7. 65 Table 4. Effects of chloroplast DNA synthesis inhibitors on DNA content of protoplasts isolated from 2 X 4 day suspension cultures of £. nigrum as determined by dot blot hybridizations. Protoplasts were isolated from 2 X 4 day cultures and treated with various inhibitors. DNA extracted from these protoplasts was applied to nitrocellulose and probed with chloroplast (psbA) and nuclear (ssRubisco) probes. Abundance of chloroplast and nuclear DNA was determined by scanning densitometry of autoradiographies. The experiments were done in triplicates.

Inhibitor Concentration Nuclear DNA cpDNA

uM % of control

None 100 100 Ethidium bromide 4.0 102 85 5-fluorodeoxyuridine 0.3 92 85 Nalidixic acid 100.0 90 65 Novobiocin 40.0 97 55 Rifampicin 10 ug/ml 89 90 66 nalidixic acid treatment had little effect on nuclear DNA content/cell. Based on these results, and those obtained from growth and recovery studies, it appears that nalidixic acid and novobiocin treatment were the most effective in selectively reducing chloroplast DNA content/cell without reducing cell viability.

2.4 DXSCOBBION Ideally, a specific inhibitor of chloroplast DNA synthesis should not have any secondary or cytotoxic effects on cell growth unassociated with the reduction in chloroplastDNA content. Although it was difficult to identify secondary effects of the various inhibitors screened, we demonstrated that certain inhibitors, while reducing DNA content, also substantially reduced cell viability. The most cytotoxic of the inhibitors tested were ethidium bromide and 5-fluorodeoxyuridine. In contrast, cell suspensions treated with the DNA gyrase inhibitors, novobiocin and nalidixic acid, effectively recovered cell growth following transfer to media lacking these inhibitors. Furthermore, we observed that among those inhibitors tested only novobiocin treated suspension cultures were capable of regenerating plants from calli. Thus, among the inhibitors tested, the DNA gyrase inhibitors appeared to have the fewest of side effects on the cultures. Both chloroplast DNA and nuclear DNA content were determined on the basis of fresh weight as well as cell 67 number. However, as previously mentioned, several of the inhibitors had adverse effects on culture viability. Since several of the inhibitors appeared to be cytotoxic, determinations of nucleic acid content from dot blot signals standardized on the basis of fresh weight may give an inaccurate estimate of the biologically active DNA content. As a result, we also quantified DNA isolated from protoplasts by dot blot hybridization and expressed DNA content on the basis of cell number. Two of the inhibitors, 5-fluorodeoxyuridine and rifampicin, which were previously shown to reduce substantially chloroplast DNA and nuclear DNA content on a fresh weight basis, had no effect on either chloroplast or nuclear DNA content expressed on the basis of viable cell numbers. Since it is difficult to rationalize 50% losses of nuclear DNA content without lethality we suggest that DNA determinations based on cell number are more accurate than those based on fresh weight. Based on measurements of DNA content per cell number only two of the inhibitors substantially reduced chloroplast DNA content. These were the DNA gyrase inhibitors, nalidixic acid and novobiocin, which reduced chloroplast DNA content per cell by 35 to 45%. Inhibition of chloroplast DNA synthesis by the DNA gyrase inhibitor nalidixic acid has previously been demonstrated in Euglena (Lyman et al 1974, Pienkos et al 1974). Nalidixic acid treatment led to irreversible bleaching of Eualena cells accompanied by a substantial decrease in chloroplast DNA (as 68 determined by density gradient fractionation of total nucleic acids). Interestingly, the chloroplast DNA content of dark- grown Euqlena or mutant strains unable to carry out photosynthetic electron transport was not reduced by nalidixic acid treatment (Lyman 1967, Lyman et al 1974). Thus, nalidixic acid was only effective in reducing Euqlena chloroplast content when cells were photosynthetically competent. In contrast, it does not appear that light is required for nalidixic acid dependent chloroplast DNA content reduction in £. nigrum suspension cells (nongreen, dark grown). However, nalidixic acid and/or novobiocin treatment of £. nigrum suspension cultures did reduce culture growth by 50%. The fact that both nalidixic acid and novobiocin reduced culture growth suggests that reductions in chloroplast DNA content in suspension cells leads to reduced culture vigor. One possible outcome of chloroplast DNA reductions could be a reduction in plastid numbers per cell (Possingham and Lawence, 1983) . While we have not determined whether plastid numbers are reduced by these treatments it is conceivable that reductions in plastid numbers could lead to reductions in the nonphotosynthetic metabolic activities compartmentalized in plastids (e.g. lipid, amino acid, and terpenoid synthesis) and, therefore, a reduction in culture growth. The mechanism by which chloroplast DNA content/cell is reduced by DNA gyrase inhibitors is most likely due to inhibition of chloroplast DNA synthesis rather than by 69 acceleration of chloroplast DNA degradation. In support of this hypothesis, we note that Weissbach and coworkers (Heinhorst et al., 1986) have shown that chloroplast DNA synthesis is inhibited in nalidixic acid treated suspension cultures (tobacco and soybean) as determined by the level of [3H]thymidine incorporation into chloroplast DNA. We found that, in addition to nalidixic acid, the DNA gyrase inhibitor novobiocin also preferentially reduced chloroplast DNA content/cell. Both of these inhibitors have been shown to reduce DNA synthesis levels in bacteria (Lam and Chua, 1987; Sancar and Sancar, 1988; Thoms and Wackernagel, 1987). Recently, a novobiocin sensitive chloroplast DNA gyrase has been partially purified from which can alter the superhelical density of cloned chloroplast DNA sequences (Lam and Chua, 1987). Alterations in the superhelical density of cloned chloroplast genes has been shown to affect their transcription rates in vitro (Lam and Chua, 1987; Stirdivant et al., 1985). However, it is unlikely that inhibition of chloroplast DNA transcriptional activity by DNA gyrase inhibitors causes reductions in chloroplast DNA content/cell, since rifampicin, an inhibitor of chloroplast RNA polymerase, had no effect on cpDNA content. We suspect that, similar to gyrase catalyzed replication processes in bacteria, the cpDNA gyrase may be required for removal of DNA supercoils introduced during DNA replication or for decatenation of replicated chloroplast DNA circles (Lam and Chua, 1987; Sancar 70 and Sancar, X988; Thoms and Wackernagel, 1987). Last of all, we note that treatments which reduce or inhibit DNA synthesis in bacterial cells such as novobiocin and nalidixic acid treatment have been shown to enhance DNA recombinational and mutational processes (Cox and Lehman, 1987; Sancar and Sancar, 1988; Thoms and Wackernagel, 1987). In fact, recA protein synthesis is promoted by nalidixic acid treatment of Escherichia coli (Thoms and Wackernagel, 1987). These observations suggest that nalidixic acid and novobiocin treatment of plant cells may also affect chloroplast DNA recombinational processes. In support of this hypothesis it is noted that Boynton et al (1990) found that treatment of Ch1amvdomonas cells with 0.5 mM 5-fluorodeoxyuridine increased the frequency of chloroplast DNA transformation (by homologous recombination) by 20- to 280-fold while also selectively lowering the chloroplast DNA content (Boynton et al., 1990). Preliminary results obtained from our laboratory indicate that novobiocin treatment of Ch1amydomona s cells also reduces chloroplast DNA content as well as enhances chloroplast DNA transformation frequencies by homologous recombination. These results suggest that transformation of higher plant chloroplast DNA by recombination mechanisms may be facilitated by treatment of cells with DNA gyrase inhibitors. CHAPTER III

CHLOROPLAST TRANSFORMATION AND MUTAGENESIS

3.1 INTRODUCTION Various genetic transformation techniques have been rapidly developed in plant genetic engineering over the past 10 years. The technique most frequently applied makes use of a natural plant-directed gene vector, the soil bacterium Aarobacterium tumefaciens (Willmitzer, 1988). In this system genes, which are inserted between the 25 direct repeats of the T-DNA region of the Agrobacterium Ti plasmid are mobilized and integrated into the plant . A second gene transfer system is the direct transfer of DNA into plant protoplasts by either, electroporation (Fromm et al., 1986), polyethylene glycol-stimulated direct DNA uptake (Lorz et al., 1985) or microinjection (Crossway et al., 1985). However, the introduction of specific genes into plants by the above techniques is not without limitations. The utilization of Ti plasmid mediated transformation in many plants has proven difficult, primarily due to the inability to efficiently regenerate plants from single cells. Direct gene transfer techniques also require the enzymatic removal of cell walls. Compared to nuclear transformation, however, chloroplast genetic manipulation is relatively unexplored and

71 72 above transformation methods have not been extended to enable reproducible transfer of foreign DNA into the chloroplast. Obviously, the genetic transformations of chloroplasts would have significant agronomic as well as basic research value. Many economically important gene products are chloroplast encoded, such as the protein conferring atrazine sensitivity or resistance, and large subunit of Rubisco which catalyzes the fixation of C02 and Oz of the Calvin cycle and photorespiratary pathways. Some products encoded by nuclear genes are also functional within the chloroplast (e.g. 3-enol- pyruvoylshikimate-5-phosphate synthase, which confers sensitivity or resistance to glyphosate). The chloroplast is different from the nucleus in many respects. The double membranes of chloroplasts are closed while the has nuclear pores open to cytoplasm. The number of genomic copies present in the chloroplast is much higher than in the nucleus, ranging from 80 in the single chloroplast of Chlamydomonas reinhardtii (Harris, 1989) to 10,000 in the higher plant cells which may have 100 or more chloroplasts. Furthermore, the number of chloroplast genomes per organelle and the number of chloroplasts per cell may vary during development or in response to environmental changes (Possingham and Lawrence, 1983). Thus in order to obtain the stable transformation and expression of foreign genes in chloroplasts, the method of delivering the transforming DNA into the chloroplast genome, 73 the selection system for transformants, the replication of DNA in the organelle as well as copy number effects must be considered. One approach which has been used to transform chloroplasts is protoplast fusion. After the fusion, it is possible to obtain hybrids containing chloroplasts from both parents or a cybrid having foreign chloroplasts. Medgyesy et al. (1985) demonstrated chloroplast DNA recombination in higher plants via selection for regenerated plants following protoplast fusions of cells containing different chloroplast encoded antibiotic resistance markers. However, the observed frequency of chloroplast recombination events was very low. This probably was due to the large copy number of chloroplast genomes per cell and the attendant segregation problems. In addition, how chloroplast DNAs from two parental chloroplasts recombined after fusion is unclear. Recently, a new transformation technique has been developed, relying upon bombardment of recipient cells with high-velocity tungsten microprojectiles coated with foreign DNA (Klein et al., 1987). Using this delivery system, Boynton et al. (1988) and Blowers et al. (1989) have demonstrated stable transformation of chloroplasts of Chlamvdomonas reinhardtii. Transient expression of the chloramphenicol acetyltrasferase (cat) gene delivered by high-velocity microprojectiles into chloroplasts of cultured tobacco cells has also been reported (Daniel1 et al., 1990), suggesting that chloroplast transformation of 74 higher plants can be achieved by bombardment. The goal of the present study is to develop a stable chloroplast DNA transformation system. It was proposed that lowering the chloroplast genome copy number will enhance the frequency of recombination or integration of DNA into the chloroplast genome due to stimulation of DNA recombination and DNA repair systems. In addition, lowered chloroplast DNA copy numbers will facilitate the segregation of transformed or mutagenized chloroplasts from wild type chloroplasts.

3.2 MATERIALS AND METHODS 3.2.1 Plasmids and Chlarovdomonas strains Ch1amvdomonas reinhardtli mutant strain cc-744 was used in this work. The cc-744 mutant was isolated following mutagenesis with 5-fluorodeoxyuridine and x-ray irradiation according to Myers et al (1982). This mutant has two 9 kb deletions spanning both psbA genes. The strain is non- phototrophic because of photosystem II deficiency. The strain was grown on Tris-acetate-phosphate (TAP) on a rotary shaker at room temperature under illumination. Plasmid p50 was constructed by subcloning the chloroplast DNA Bam HI fragment from wild type Chiamvdomonas reinhardtii in pUC 8 (Harris et al.f 1987). The insert in pUC 8 is about 20 kb long and carries the entire psbA gene. Plasmid pPPX was constructed by subcloning a 0.77 kb Pst I/Xba I fragment of psbA gene which confers resistance to 75 atrazine in Amaranthus hybridus (Hirschberg and McIntosh 1983) . Plasmid pAH484 was constructed by subcloning cpDNA EcoR 1 fragment from atrazine resistant h±. hybridus in pBR322 . The insert is 3.68 kb long and contains a complete psbA gene. Plasmid pSSUl60 was constructed by subcloning a genomic Hind III fragment from Pisum sativum in pBR322. The insert is 0.8 kb long and encodes the small subunit of chloroplast ribulose- 1,5-bisphosphate carboxylase (Bedbrook et al., 1980).

3.2.2 Reduction of Chloroplast DNA Content in Chlamvdomonas Cells of the non-photosynthetic mutant strain cc-744 were grown in TAP medium, containing one of the following potential cpDNA synthesis inhibitors: 0.5 mM 5-fluorodeoxyuridine, 100 mM novobiocin or 10 ug/ml rifampicin for two days. Control cells were grown in TAP medium without cpDNA synthesis inhibitors. The DNA from same number (1 X 109) of cells of different treatments was isolated according to mini-prep procedure developed by Dellaporta et al. (1983). The chloroplast and nuclear DNA contents were analyzed using the dot blot hybridization procedure as described in Chapter II. The 3 kb Xba I DNA fragment encoding the psbA gene from pX13 was isolated and labeled by the random primer extension method (Feinberg and Vogel stein, 1983) and used as a chloroplast DNA probe. The fragment encoding the 18S ribosomal RNA was used as a nuclear DNA probe. Known amounts of chloroplast and nuclear DNA standards corresponding to 10, 30, 100, 300 and 1,000 ng 76 of plasmid X-13 containing the psbA gene (chloroplast marker) and plasmid pSSU-160 containing the nuclear encoded fragment were immobilized on the same membrane in order to confirm the linearity of dot blot signals.

3.2.3 Chlamvdomonas Transformation Cells of the nonphotosynthetic mutant strain cc-744 were harvested at a density of 2 X 106 cells after treatment with chloroplast DNA synthesis inhibitors to reduce the chloroplast DNA copy number. Aliquots (1.0 ml) containing 2 X 107 cells were plated onto Petri dishes of TAP media and swirled uniformly over the surface of the media. To absorb DNA to the microprojectiles, 2.5 ul of p50 (at 1 ug/ul) was added to 25 ul of a suspension of 1.2 um tungsten particles (60 mg/ml of 50% glycerol) in a 1.5 ml microfuge tube. After addition of the DNA, 25 ul of 2.5 M CaCl2 and 10 ul 0.1 H spermidine free base were added to the suspension. After 10 min of incubation, the particles were pelleted by centrifugation in a Microfuge for one min and 45 ul of the supernatant was removed to concentrate the particles for 3 bombardments. The clumps of particles were dispersed by brief sonication and 5 ul of DNA coated tungsten particles were placed on the front surface of a polyethylene macroprojectile. The macroprojectile was then placed into the barrel of the particle gun. P50 coated particles were fired onto a plate of cells at a distance of 15 cm. Bombardered cells were resuspended in 1 ml of HS media 77 (Sueoka 1960) and spread onto a plate of minimal HS media and incubated at room temperature under bright light. Only cells which are capable of photosynthesis, i.e. transformed cells, can grow on HS media. Photosynthetically competent cells were counted as green colonies from six transformation experiments.

3.2.4 Regeneration of Plants from Suspension Cells and Leaves Green calli and shoots were induced from suspension cells of S. nigrum grown on the MS medium containing 1% sucrose, 2.5 uM 6-benzylaminopurine (6-BAP) and 10% coconut milk. The medium containing MS salt, Nitsch vitamins (Nitsch and Nitsch, 1969), 1% sucrose, 0.5 mg/L 6-BAP and 0.05 mg/L indoleacetic acid (IAA), was used to induce shoots from young leaves of £. niarum. After formation of a shoot, roots were induced by growing plants in MS medium without plant hormones. The callus and leaves of §. niarum were also cultured on the above regeneration medium containing various concentrations of atrazine. Atrazine at 60 uM was lethal to the wild type line while an atrazine resistant line grew normally even when 100 uM of atrazine was in the medium.

3.2.5 Transformation of Solanum niarum Cells Solanum niarum suspension cells were grown as described in Chapter II in the presence of 40 uM novobiocin for 4 days and then transferred to fresh culture medium plus 40 uM novobiocin for one more day before bombardment. About 500 mg 78 of suspension cells (fresh weight) were collected by centrifugation and placed on one layer of filter paper. The single filter paper bearing the cells was then placed over shoot regenerating medium containing 60 uH atrazine in a petri dish. To prepare £. nigrum leaf cells for bombardment with microprojectiles, young leaves from 14-20 days germinated seedings, grown in the MS medium containing 40 uM novobiocin, were collected and placed directly on the plate containing selection medium. The DNA used for transformation was pPPX, containing a 0.77 Kb Pst 1/Xba insert of the psbA gene fragment which confers resistance to atrazine in Amaranthus hvbridus (Hirschberg and McIntosh, 1983). The samples were bombardered with 0.5 uM tungsten particles coated with DNA. The DNA-coated particles and bombardment procedure were the same as described for Chlamydomonas transformation. The bombarded cells were maintained at 25°C in the light in a plant growth chamber. The green shoots that developed on shooting regeneration medium plus 60 uM atrazine were transferred to MS medium plate plus 100 uM atrazine to regenerate roots.

3.2.6 Southern Hybridization Analyses DNA was isolated from leaves of plants regenerated from bombarded cells using the mini-prep procedure (Dellaporta 1983) . For DNA gel blot hybridizations, DNA was restricted with Tag I and electrophoresed through 1.6% TBE-agarose gel 79 (Maniatis et al., 1982), and transferred to a nitrocellulose membrane by the method of Southern (1975). The chloroplast DNA probe (3 kbp Xbal DNA fragment encoding psbA from pX13) was isolated and labeled by the random primer extension method (Feinberg and Vogelstein, 1983) . The hybridization was carried out according to the procedure described in Chapter II.

3.2.7 Preparation of Oligonucleotide Probe In order to determine changes of nucleotide sequence in selected atrazine resistant plants two oligonucleotide probes constructed using a Beckman DNA synthesizer were used in this study. The synthetic oligonucleotide complementary to the sequence found in the atrazine susceptible Solanum niarum is 5'ATATGCTAGTTTCAA3* (probe designated "psbAA"), whereas the oligonucleotide complementary to the sequence found in atrazine resistant Amaranthus hybridus is 5'ATATGCTGGTTTCAA3' (probe designated "psbAG"). There is a single base pair difference between psbAA and psbAG at the 8th nucleotide from 5* end. The were end-labelled with 32P using the T4 kinase protocol as follows: The oligonucleotide were heated in a tightly sealed glass vial at 55°C for 12 hours then dried down in a speed vacuum. The dried oligonucleotides were resuspended in 150 ul TE, 30 ul 7.5 M ammonium acetate and 750 ul ethanol and precipitated at -2 0°C overnight. The oligonucleotides were then pelleted at 13,000 rpm at 4°C in a microfuge, washed once with 70% ethanol and resuspended in 150 80 ul of TE. For labeling, 100 ug of oligonucleotide was added to 25 ul of a reaction mixture containing IX kinase buffer (70 mM Tris/HCl, pH 7.5, 10 mM MgCl2, 100 Mm Kcl, 5 mM DTT) , 100 uCi r-32P-ATP and 10 units of T4 kinase and incubated at 37°C for 40 min. The reaction was stopped by adding 25 ul 7.5 M ammonium acetate and 5 volumes of ethanol. After precipitation at —70°C for 15 min, the oligonucleotide was pelleted at 13,000 rpm in a microfuge and resuspended in 100 ul TE. Following addition of 50 ul 7.5 M ammonium acetate and 5 volumes of ethanol, the oligonucleotides were re-precipitated and washed with 70% ethanol as above. The labelled oligonucleotides were dried in a vacuum and resuspended in TE buffer.

3.2.8 DNA Hybridizations Dot blots of genomic DNA or plasmid DNA were prepared as follows: 100 ul of 1 M NaOH was added to 100 ul of diluted DNA samples and the DNA was allowed to denature at room temperature for 20 min; 200 ul of 2 M ammonium acetate was then added and aliquots were vacuum-suctioned in a dot-blot apparatus onto nitrocellulose filters that were pre-wetted with water followed by 1 M ammonium acetate. The membranes were baked for 2 hours at 80°C in a vacuum oven. The filters were then pre-washed in a solution containing 50 mM Tris, pH 8.0, 1 M NaCl, 1 mM EDTA and 0.1% SDS at 37°C for one hour. Prehybridizations were carried out in 5X SSPE, 0.1% SDS, 5% 81 dry milk and 100 ug/ml calf thymus DNA at 37°C overnight. The oligonucleotide probes (IX 107 cpm) were added to the bag of prehybridization solution and incubated at 37° for 20 hours. All the filters were then washed twice at 24°C in 6X SSC for 15 min. The melting temperature (Tm) of the hybrid was determined according to the method of Meinkoth and Wahl (1984). Three filters probed with same labeled oligonucleotides were washed two times at 34°C, 37°C and 40°C separately for 20 min each wash. The filters were then exposed to X-ray film at -80°C using intensification screens.

3.2.9. Chloroplast Mutagenesis The seeds of wild type niarum were germinated on MS medium with or without 40 uM novobiocin. After two weeks growth, the young leaves of seedlings were transferred onto the regeneration medium as decribed before plus 40 uM novobiocin and 60 uM atrazine. Leaves growing on the same medium in the absence of novobiocin were used as a control. The regenerated shoots were then transferred onto MS medium containing 100 uM atrazine to regenerate roots.

3.3 RESULTS 3.3.1 Regulation of Chloroplast DNA Content in Chlamvdomonas To test the toxicity of novobiocin and rifampicin, Chlamvdomonas cells were grown in the presence of different amounts of these compounds. The results have shown that the 82 alga can grow in the presence of 100 uM novobiocin or 10 ug/ml rifampicin without a significant reduction in cell growth and viability. Chlamvdomonas wild type strain CC-125 were treated with 5-fluorodeoxyuridine (0.5 mM) , novobiocin (100 urn) as well as rifampicin (10 ug/ml) to determine the effects of these inhibitors on cpDNA synthesis. To determine nuclear and chloroplast DNA content in inhibitor-treated cells, total DNA extracted from same number of cells was blotted onto nitrocellulose and hybridized with a probe prepared from a chloroplast encoded gene (psbA) as well as the gene encoding the small subunit of Rubisco (nuclear encoded). Signals obtained upon autoradiography of the dot blots were quantified by scanning densitometry. As shown in the Table 5, novobiocin effectively reduced chloroplast DNA to 60% of the level in the untreated control. In contrast, the amount of nuclear DNA, present in novobiocin treated cells was largely unaffected. The content of chloroplast DNA per cell was also reduced by 5- fluorodeoxyuridine to 69% of the untreated control, whereas the level of nuclear DNA was not changed. These results indicate that nuclear DNA synthesis is not impaired by exogenous addition of the thymidine analogue, Fudr. Nuclear and chloroplast DNA contents was not significantly affected by rifampicin, indicating that rifampicin is not an inhibitor of chloroplast DNA synthesis in Chlamvdomonas♦ 83 Table 5. Effects of cpDNA synthesis inhibitors on DNA content of Ch1amvdomonas as determined by dot blot hybridizations. Chlamvdomonas were cultured in TAP medium with cpDNA synthesis inhibitors for 5 generation. DNA extracted from Ch1amvdomonas cultures with various inhibitor treatments was applied to nitrocellulose and probed with chloroplast (psbA) and nuclear (ssRubisco) probes. Abundance of chloroplast and nuclear DNA was determined by scanning densitometry. The experiments were done in triplicate.

Inhibitor Concentration Nuclear DNA cpDNA (uM) % of Control

None -- 100 ± 3 . 7 100 ± 4.1 5-fluorodeoxyuridine 500 96 ± 5.4 68 ± 3.9 Novobiocin 100 95 ± 3.3 60 ± 4.7 Rifampicin 10 ug/ml 98 ± 4.9 86 +5.8 84 3.3.2 Transformation Frequency in Chlamvdomonas Using the particle bombardment methods, the inhibitor treated, photosynthetically defective mutants, were transformed with cloned wild type chloroplast psbA gene in plasmid p50 and screened for restoration of photosynthetic capacity. The Ch1amvdomonas reinhardtii strain (CC-744) carrying a 9 kb deletion in the psbA gene encoding the D1 protein of the PSII complex was used as the recipient cell for transformation experiments. To reduce chloroplast genome copy number, the CC-744 cells were grown in the presence of 0.5 mM 5-fluorodeoxyuridine or 100 uM novobiocin. The cells were then bombarded with p50 plasmid carrying the 20 kb Bam HI chloroplast restriction fragment containing the entire wild type psbA gene. After transformation, cells were transferred to high salt medium (HS) which has no acetate. Since CC-744 cells are defective in photosynthesis, they can not grow on HS medium. Only cells expressing the psbA gene product can survive on HS selection medium. Thus, green colonies growing on HS medium were considered as transformants. One week after transformation, photosynthetically competent green colonies formed on the selection medium, indicating that the wild type gene and gene product are present in these cells. The restored wild type psbA function persisted in all transformants for at least one month of culture on the selection medium. These results apparently showed that the psbA gene deletion was restored to a functional size by transformation with wild type 85 psbA gene. Pretreatment of Ch1amvdomonas cells with chloroplast DNA synthesis inhibitors was shown to increase the frequency of transformation. As shown in Table 6, pregrowth of the deletion mutants in 5-fluorodeoxyuridine or novobiocin increased transformation frequency 3 5 and 24 fold respectively. These results suggest that transformation of plant chloroplast DNA by recombination mechanisms may be facilitated by treatment of cells with DNA gyrase inhibitor.

3.3.3 Regeneration of Plants from Suspension and Leaf Cells Suspension cells of Solanum niarum were chosen as the target to transform chloroplasts of higher plants in our studies, based on two considerations: chloroplast genome copy number in £. niarum can be reduced by chloroplast DNA synthesis inhibitors as discussed in Chapter II; and atrazine resistant psbA biotypes are available for use as chloroplast selectable markers. It is important to determine the favorable culture conditions for plant regeneration from suspension cells. (e.g. IAA) and (e.g. BAP) are two types of phytohormones needed in culture and their relative concentrations in the medium often control the pattern of differentiation. To determine the concentration of BAP and IAA to be used for the plant regeneration, various combinations of these two phytohormones were tested. After two weeks, cell Table 6. Increase of the transformation frequency by pre-treatment of

Chlamvdomonas with cpDNA synthesis inhibitors. Using the particle bombardment process, photosynthetically defective mutants of

Chlamvdomonas reinhardtii were transformed with the p50 plasmid carrying the 20 kb BamHl chloroplast restriction fragment containing the wild

type psbA gene. Total colonies which recovered their photosynthetic

capacity were counted from six transformation experiments.

Estimated Treatment No. of Transformants Transformant Frequency

None 2 0.2 X 10'8

5-fluorodeoxyuridine 70 5.8 X 10'8

Novobiocin 48 4.0 X 10'8

OB o\ 87 colonies grew rapidly to form an intensely green, compact callus under continuous illumination. Cells within cell clumps or callus developed into globular meristematic cell masses. From some green calli, shoot buds developed. The highest frequency of inducing green shoots was obtained from medium containing 0.5 mg/L BAP in combination with 0.05 mg/L IAA (Table 7). Shoot differentiation was also induced from young leaves of S. niarum by using the same shoot regeneration medium. Alternatively, similar results were obtained using MS medium containing 1% sucrose, 2.5 uM BAP and 10% coconut milk.

3.3.4 Chloroplast Transformation and Mutagenesis Wild type Solanum niarum suspension cultures and seedings were treated with the DNA gyrase inhibitor, novobiocin (40 uM) to reduce chloroplast genome copy number. The calli and leaves were then transformed with a plasmid pPPX containing a 0.77 kb Pst I/Xba I insert of the psbA gene fragment which confers resistance to atrazine in Amaranthus hvbridus (Hirschberg and McIntosh 1983) . For selection of transformants all bombarded cells were plated on the shoot regeneration medium containing 60 uM atrazine. Small green calli resistant to atrazine were isolated and grown on the same shoot regeneration medium containing 100 uM atrazine. After two months, calli with regenerated shoot were moved to MS medium containing 100 uM atrazine. Table 7. Effect of BAP and IAA on green shoot formation from callus. The number shown in the Table is the number of green buds formed from 100 calli derived from suspension cultures.

BAP IAA 0.1 mg/1 0.5 mg/1 0.l mg/1

0.01 mg/1 29 24 18

0.05 mg/1 17 41 28

0.1 mg/1 2 21 27

09 09 89 After regeneration of plants under atrazine resistance selection, potential transformants were screened by Southern blot analysis for restriction fragment length polymorphism. The psbA atrazine resistance gene from &. hvbridus has a unique restriction endonuclease Taq 1 site (Fig 8) . Since this site is only 22 base far from the site of a single A to G base transition accounting for change from atrazine-susceptible to atrazine-resistant in both hvbridus and S. nigrum, these two sites have a high probability to be transferred together during homologous recombination. In transformed chloroplasts it is expected that the psbA gene of Amaranthus hvbridus integrated into the chloroplast genome of £. nigrum by homologous recombination generating either a if a single cross over event occurs or a gene replacement if it occurs by a double cross over. Alternatively, the DNA may randomly integrate into the genome. These possibilities were tested by Southern blot analysis of the chloroplast DNA from selected atrazine resistant plants with psbA gene probe. In the case of random integration, the extra Taq 1 band of 270 bp will be seen in the cpDNA of transformants (Fig. 8C). If the psbA genes in recipient chloroplasts were replaced by homologous recombination, two Taq 1 fragments (940 and 270 bp) are expected to be produced (Fig. 8C) . Unfortunately, none of these plant were true transformants since these two expected bands were not observed by Southern hybridization analyses (Fig. 9). 90

Figure 8. Restriction naps o 1 Tag 1 fragment of psbA gene in

A. hvbridus (A) , £. nigrum (B) and possible transformants (C) or (D) . o, Taq 1 site; □, the site of a single A to G base transition accounting for change from atrazine-susceptible to atrazine-resistant in both A. hvbridus and fi. nigrum; the site responsible for atrazine-susceptible. 91

1 2 3 4 5 6 7

1,500 b

1,250 b

1,000 b 700 b 500 b

Figure 9. Southern blot analyses of selected atrazine resistant plants. DNA was prepared from leaves of wild type S. nigrum (lane 1) , atrazine resistant S. nigrum (lane 2), and plants regenerated from bombardered cells (lane 3-7). DNA samples were digested with Taq 1, separated by 1.6% agarose gel in TBE buffer and blotted onto nitrocellulose before hybridization with a cloned EcoR 1 fragment of psbA gene from plasmid pAH484. *, The expected size of Taq 1 fragments in transformants. 92 Although the plants tested were not transformed, they were resistant to atrazine. One possible explanation for this is that inhibitor treatment induced mutations in the cpDNA which resulted in atrazine-resistance. To test this possibility, the cpDNA isolated from selected atrazine resistant plants was also probed with two synthetic oligonucleotides. One oligonucleotide, psbAA (5'ATATGCTAGTTTCAA3') is complementary to the sequence found in atrazine susceptible S. nigrum. The other, psbAG (5'ATATGCTGGTTTCAA3') is complementary to the sequence in atrazine resistant hvbridus. Different temperatures for washing the hybridization membrane were used to determine the binding stringency between the oligonucleotide probe and the chloroplast DNA sequences to be tested. As the wash temperature was increased the signals of plasmid DNA containing atrazine susceptible psbA gene decreased when probed with the psbAG probe while the signals did not decrease when probed with the complementary psbAA probe (Fig. 10). In contrast, the plasmid DNA containing atrazine resistant psbA gene showed the opposite results, i.e. the psbAG probed hybridized more strongly. Chloroplast DNAs isolated from selected atrazine resistant plants, e.g. plant 1 and plant 2 had a higher affinity for the psbAG oligonucleotide, which has an single base pair change which accounts for atrazine resistance, than chloroplast DNA from wild type. These results suggest that cpDNA synthesis inhibitors may have mutagenic Figure 10. Wash stringency for selected atrazine resistant plants. Genomic DNA and plasmid DNA were blotted on the nitrocellulose and hybridized with oligonucleotide probes. Three filters probed with the same labeled oligonucleotide were washed two times at 34°C, 37°C and 40°C separately. The filters were then exposed to x-ray film at -80°C using intensification screens. A: 1,000, 500, 250, and 125 ng of plasmid pMLS containing the atrazine susceptible psbA gene (Bam HI fragment) were loaded on slots 2-5; 1,000, 500, 250 and 125 ng of plasmids pAH484 containing the atrazine resistant psbA gene were loaded on slots 8-11. B: 2 ug and 1 ug of ij£. niarum cpDNA isolated from atrazine susceptible wild type plant were loaded on slots 1-2; 2 ug and 1 ug of £. niarum DNA isolated from atrazine resistant regenerated plants were loaded on the slots 7-8. C: 2 ug and 1 ug of chloroplast DNA isolated from selected atrazine resistant plant were also loaded on the slots 1-2 (plant 1) ; 5-6 (plant 2) and 9-10 (plant 3). Slot 13-24 is the repeat of the slot 1-12. DNA in slots 1-12 was probed with psbAA complementary to the sequence found in atrazine susceptible S. niarum. DNA in slots 13-23 was probed with psbAG complementary to the sequence found in atrazine resistant &. hvbridus.

93 a\

34°C 37°C 40 °C 0 I I 00 a O m | | I i il II' g f till i l Itill II ' I I I IIII t , > u) (h o r-t o h ( ) u r>. O l v -, I 1 I ' I I I I I I cm o i > * C<' n N ^ O C0 ui ro — (J CM M C (SJ r—4 I I I I i i 11 I I I II II i II li I II 1 I I I I I i li 11 II I I IIM

Figure 10. 95 effects on the chloroplast DNA genome. To test this hypothesis, the seeds of wild type £. niarum were germinated on MS medium with or without 40 uM novobiocin. After two weeks growth, the young leaves of novobiocin treated seedlings were transferred onto the shoot regeneration medium containing 4 0 uM novobiocin and 60 uM atrazine. As a control, leaves of non-treated seedlings were grown on the same shoot regeneration medium plus only 60 uM atrazine. The regenerated shoots were then transferred onto MS medium containing 100 uM atrazine to regenerate roots. As shown in Table 8, 43 atrazine resistant plants were selected from 10,000 novobiocin-treated seeds whereas only 2 atrazine resistant plants were selected from non-treated seeds. This 2 0 fold increase in the frequency of mutagenesis by novobiocin treatment suggests a mutagenesis effect of this gyrase inhibitor on the chloroplast DNA.

3.4.DISCUSSION Our studies on the regulation of chloroplast DNA content have demonstrated that chloroplast genome copy number in Chlamydomonas reinhardtii can be selectively reduced by treatment with chloroplast DNA synthesis inhibitors. The results of 5-fluorodeoxyuridine treatment are consistent with previous studies (Wurtz et al., 1977) which indicate that growth of Chlamvdomonas in the presence of 5- fluorodeoxyuridine leads to a reduction in chloroplast DNA content, whereas nuclear DNA levels and growth rate were 96

Table 8. Mutagenic Effect of Novobiocin on Chloroplast DNA. The seeds of wild type §_. niarum were germinated on MS medium with or without 40 uM novobiocin. After two weeks growth, the young leaves of seedlings were transferred onto the regeneration medium as decribed in materials and methods plus 40 uM novobiocin and 60 uM atrazine. Leaves growing on the same medium in the absence of novobiocin were used as a control. The regenerated shoots were then transferred onto MS medium containing 100 uM atrazine and the regenerated plants were counted as atrazine resistant plants.

Treatment Total Seeds Atrazine Resistant Plants

Non-treated 10,000 2 Novobiocin treated 10,000 43 97 largely unaffected. As described in Chapter II, the two DNA gyrase inhibitors, nalidixic acid and novobiocin, preferentially reduced cpDNA levels in S. niarum with little effect on the nuclear DNA content. Similarly, novobiocin is also effective in lowering copy number of chloroplasts in Chlamvdomonas without reduction in cell growth and viability. It has been reported that bacterial cells treated with nalidixic acid have higher DNA recombination and mutation frequencies (Thomas and Wackernagel, 1987). In Chlamvdomonas. pretreatment with 5-fluorodeoxyuridine also increased the frequency of chloroplast DNA transformation (Boyton et al., 1988). These observations suggest that novobiocin not only inhibits chloroplast DNA replication but may also affect chloroplast DNA recombination processes in Chlamydomonas. It was predicted that chloroplast transformation frequency would be increased by treatments which reduced the number of chloroplast genomes prior to transformation. Results of the present study have verified this prediction. When Chlamydomonas mutant strain CC-744 were grown in the presence of 5-fluorodeoxyuridine, the frequency of chloroplast transformation increased by 35 fold. Similar effects of fluorodeoxyuridine were also reported when different Chlamydomonas mutants were used (Boynton et al., 1990; Kindle et al., 1991). Using the same cell strain, novobiocin treatment increased chloroplast transformation by 24 fold. In the present studies, the growth rate of CC-744 cells was not 98 affected by the treatment with 0.5 mM fluorodeoxyuridine and 100 uM novobiocin. In fact, the reduction in copy number of chloroplast genes in wild type cells by fluorodeoxyuridine has little or no effect on the amounts of photosynthetic polypeptides synthesized (Hosier et al., 1989). Thus the reduction in chloroplast genome copy number is likely to increase the probability for donor genes to be expressed in recipient chloroplasts and to facilitate segregation of transformants from non-transformed cells. Our results have confirmed that it is possible to introduce modified genes into the chloroplasts of Chlamvdomonas cells to assess their function in vivo. The ability to genetically transform this single-celled eukaryotic alga greatly increases their experimental research value. First, various factors affecting bombardment transformation of chloroplasts could be studied in detail in Chlamydomonas. providing a model system for transformation of chloroplasts in higher plants. Secondly, bombardment transformation could be used to isolate and characterize chloroplast genes with known phenotypic effects. Third, transformation combined with mutagenesis techniques would provide a new way to study photosynthesis in chloroplasts. Fourth, it will be possible to investigate recombination and DNA repair mechanisms in chloroplasts. Finally, genetic manipulation and transformation in both the nucleus and chloroplasts should make Chlamvdomonas an ideal organism for studies of chloroplast-nuclear 99 interaction. Using a stable nonphotosynthetic deletion mutant lacking the chloroplast gene psbA encoding the D1 reaction center polypeptide of photosystem II as a DNA recipient combined with the corresponding cloned wild type gene as a donor provides an efficient transformation system. This resulted in complimention of the mutant phenotype under selection pressure and photosynthetically competent cells were readily detected even at very low transformation frequencies. Boynton et al (1989) have also reported successful transformation using atpB gene deletion mutants as recipient cells. Therefore, the deletion of chloroplast genes encoding proteins required for photosynthesis provides a good selection marker for chloroplast transformation. In addition, such deletions may also be used for co-transformation with complementary sequences along with a non-selectable gene manipulated in vitro. Since transformants can only survive with an integrated gene, unselected DNA sequences will likely be co-introduced into the recipient genome. Thus the effects of this mutagenesis can be determined, even in the absence of direct selection (Roffey et al., 1991). An ideal chloroplast transformation system would permit DNA integration into all chloroplast genome copies by homologous recombination with the wild type gene copy. Unfortunately, evidence for the occurrence of genetic recombination in higher plant chloroplasts is rather limited. 100 Protoplast: fusion Is one method that has been used for higher plant chloroplast transformation. After fusion, it is possible to obtain hybrids containing chloroplasts from both parents, or cybrids that have foreign chloroplasts (Binding et al., 1982; Pelletier et al., 1983). It has been shown that the genetic recombination between chloroplasts in fused can occur (Gleba et al., 1985). Maliga and coworkers demonstrated chloroplast DNA recombination in higher plants via selection for regenerated plants from protoplast fusions of cells containing different chloroplast-encoded antibiotic resistance markers (Medgyesy et al., 1985; Svab et al., 1990). Recently another method, bombardment transformation, has been developed and can be used reproducibly to transform chloroplasts in Chlamvdomonas (Boynton et al., 1988; Blowers et al., 1989) . Bombardment transformation has several advantages over with protoplast fusion. First, making protoplasts and regenerating plants from protoplasts is not required. By using high- velocity microprojectiles, foreign DNA can be introduced into intact cells and tissues. Thus it would be broadly applicable. Second, it is simple and effective. Large numbers of cells can be transformed within a short time. Third, it is easier to manipulate the transformation donor as a piece of DNA rather than as an intact protoplast or chloroplast. Fourth, the bombardment method allows only a selected DNA sequence to be introduced into the host cell or chloroplast, whereas 101 protoplast fusion mixes the genomes of two different cells. Another problem associated with protoplast fusion is that chloroplasts from two different parent cells may segregate, generating somaclonal variation. Each chloroplast contains multiple copies of the genome, with copy numbers from 8-200 per chloroplast reported for different species (Herman and Possingham, 1980). Thus, an introduced gene will probably be diluted or lost. The hybrid cytoplasm generated by protoplast fusion usually sorts out to a stably inherited genotype after a number of cell divisions. The somatic hybrid plant usually has one or the other parental chloroplast genome (Belliard et al., 1978; Pelletier et al., 1983; Bonnett and Glimelius, 1983). Therefore, it is necessary to find a way to integrate the introduced chloroplast gene into higher percentage of host chloroplast DNA genomes. This could be accomplished by lowering the total number of chloroplasts and chloroplast DNA per cell. Using bombardment transformation techniques combined with reduction in chloroplast genome copy number, chloroplast transformation was successful in Ch1amvdomonas but not in higher plant £. niarum. To account for this, two differences between chloroplasts from these two plants may be considered. Each Ch1amydomonas cell contains a single large chloroplast adjacent to the plasma membrane along most of the periphery of the cell. In contrast, chloroplasts in higher plant cells are small in size relative to the total cell volume and lie 102 relatively far from the plasma membrane. Thus, chloroplasts in Ch1amvdomonas may provide a better target for bombardment and may have a greater chance to be transformed. Competition between recipient genes and donor genes may be another problem for chloroplast transformation in higher plants. As previously mentioned, there is more than one chloroplast and multiple copies of cpDNA in each chloroplast of higher plant cells. Since the number of chloroplast genomes in higher plants is much higher than in Chlamvdomonas. treatment with cpDNA synthesis inhibitors to reduce chloroplast DNA content may be more effective in Chlamvdomonas than in £. niarum. Thus the integrated foreign chloroplast gene must face strong competition with the resident gene for phenotypic expression. Therefore, the foreign gene is more likely to be replaced by the resident gene during segregation in higher plants. Further, since only cells on the surface of cell aggregates or calli may be transformed, the transformed cells and non- fransformed cells may form chimeras, complicating the identification or selection of rare transformation events. In addition, several physical parameters of the bombardment method (such as particle coating procedure, particle size and volume) and the physiological state of the target cells have to be taken into account. It is also important to know whether chloroplasts in higher plants have recA regulated recombination systems or not. The further testing of all factors influencing bombardment transformation in higher plant 103 cells will be necessary before high frequency gene transfer of higher plant chloroplasts will be achieved. CHAPTER IV. CHLOROPLAST RECA-LIKE PROTEIN IN CHLAMYDOMONAS INTRODUCTION Previous investigators have reported that chloroplast DNA synthesis can be selectively inhibited by a variety of compounds. In Eualena. inhibition of chloroplast DNA synthesis by nalidixic acid, a prokaryotic DNA gyrase inhibitor, has been shown to lower the steady state copy number of the chloroplast genome without affecting the nuclear DNA content (Lyman, 1967; Lyman et al., 1974? Pienkos et al., 1974). In the green alga Chlamvdomonas. treatment with the thymidine synthesis inhibitor fluorodeoxyuridine (Fudr) and/or the DNA binding dye ethidium bromide results in specific reduction in chloroplast DNA content which recovers to normal levels following removal of inhibitor (Flechtner and Sager., 1973; Heinhorst et al., 1985). As described in Chapter I and Chapter III, the two DNA gyrase inhibitors, nalidixic acid and novobiocin, have also been demonstrated to preferentially reduce chloroplast DNA steady state levels with little effect on nuclear DNA content. In addition, 5-fluorodeoxyuridine treatment of Chlamvdomonas has also been shown to increase the level of transmission of chloroplast DNA for mating type minus (mt‘) in crosses with mt+, and to increase the recovery of

104 105 chloroplast mutations (Wurtz et al., 1979). This enhancement of recombination events is similar to that observed in prokaryotic organisms when DNA synthesis is inhibited and is generally known as the "SOS" response system (Thomas and Wackernagel, 1987). In bacterial cells, treatment with novobiocin and nalidixic acid, which reduce or inhibit DNA synthesis, has been shown to enhance DNA recombination and mutation frequencies (Cox and Lehman, 1987; Sancar and Sancar, 1988; Thomas and Wackernagel, 1987). These observations suggest that nalidixic acid and novobiocin treatment of plant cells may also affect chloroplast DNA recombinational processes. In support of this hypothesis, it is noted that treatment of Chlamvdomonas with 0.5 mM 5-fluorodeoxyuridine increased the frequency of chloroplast DNA transformation (by homologous recombination) while also selectively lowering the chloroplast DNA content (Boynton et al, 1988). In our lab, we have treated chloroplast psbA gene deletion mutants of Chlamydomonas with 5-fluorodeoxyuridine as well as novobiocin to reduce number of chloroplast genomes. Using the particle bombardment method we have transformed the pretreated, photosynthetically defective mutants with cloned wild type chloroplast psbA gene and restored their photosynthetic capacity. As shown in Chapter III, pre-treatment of the deletion mutants with 5- fluorodeoxyuridine or novobiocin increased transformation frequency 35 and 24 fold, respectively. These results suggest 106 that transformation of plant chloroplast DNA by recombination mechanisms may be facilitated by treatment of cells with DNA gyrase inhibitor. Recently, evidence supporting the presence of a recombination system in chloroplasts has come from the identification of a chloroplast open which is homologous to the £. coli recA and uvr genes which are involved in recombination and DNA repair (Oppermann et al., 1989) . Unfortunately, little is known about the of chloroplast DNA recombination and repair. Our objective has been to identify and characterize the proteins which may be involved in chloroplast DNA recombination and repair. In this chapter we will present evidence indicating that: (1) a protein cross-reacting with against £. coli recA protein is present in Ch1amydomonas chloroplasts, and that this protein may be involved in DNA recombination and (2) Ch1amvdomonas chloroplast stromal extracts contain an ATP dependent recA like activity.

4.2* MATERIALS AND METHODS 4.2.1 Chlamvdomonas Strain and Plasmid Ch1amvdomonas CC-406 is a mutant strain which is cell wall-less. It was maintained in TAP liquid medium on a shaker at room temperature under continuous illumination. The cells were subcultured at 5 day intervals. 107 Plasmid p50 was constructed by subcloning a chloroplast DNA Bam HI fragment containing an intact psbA gene from wild type Chlamvdomonas reinhardtii in pUC8 (Harris et al.r 1987). The insert in pUC8 is about 20 kb long and carries entire psbA gene.

4.2.2 Isolation of Chloroplast Stromal Proteins and Western Blotting Chlamvdomonas cells were broken in a Yeda press at 80 psi in isolation buffer containing 250 mM sorbitol, 35 mM HEPES- KOH, pH 7.7, 1 mM MnCl2, 5 mM MgCl2 and 2 mM EDTA. The 3 ml of broken cells was mixed with 3 ml 80% (v/v) percoll in the isolation buffer. The above solution was then layered onto 10 ml of 60% percoll in the isolation buffer. After centrifugation at 5,000 xg for 20 min at 4°C, chloroplasts were collected between 40% and 60% percoll layer. The intact chloroplasts were then washed two times with the isolation buffer. Intact chloroplasts were broken by homogenization in the buffer containing: 20 mM Tris.HCl, pH 7.5, 1 mM dithiothreitol, 10 mM MgCl2, 50 ug/ml bovine serum albumin, 10 mM KC1 and 5% glycerol. Membrane fractions were removed by centrifugation at 100,000 X g for one hour. The chloroplast stromal fraction was solubilized in buffer containing 12.5% sucrose, 33 mM dithiothreitol (DTT) , 1.5% SDS and 0.03% Bromphenol blue for SDS-PAGE. Proteins were denatured by heating at 95°C for two minutes and separated by sodium 108 deodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (Sayre et al., 1986). Samples were loaded on the basis of equal amounts of protein (50 ug). The separated proteins were electrophoretically transferred to nitrocellulose or Immobilon-P at 250 mA for 4.5 hours in transfer buffer (25 mM Tris, 190 mM glycine, 20% (v/v) methanol, 0.1% SDS) by the western blot method of Towbin et al. (1979). Filters were blocked in 20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1.5 g/1 sodium azide (TSA) + 5% dry milk for 2 hours at 37°C and antibodies against £. coli recA provided by Dr. Roberts from Cornell University were added at dilutions of 1/200 to 1/1000. The membranes were incubated overnight at room temperature and then washed three times for ten min each in TSA. Antibody was detected in 30 ml TSA containing 5% dry milk and 1 uCi Protein A-125I for 2 hours at room temperature. Blots were washed again three times in TSA for ten minutes each and filters placed on Kodak XAR-5 film for one week.

4.2.3 Treatment of Chlamvdomonas Cells with Novobiocin Chlamvdomonas strain cw-15 was inoculated in 500 ml TAP medium and cultured for 2 days, followed by addition of 100 mM novobiocin. The same number of cells were harvested at 1, 3, 6, 9 and 12 hours after adding novobiocin. The stromal proteins were extracted, separated by SDS-PAGE and transferred to nitrocellulose, and immunodetected as described above. The abundance of proteins cross-reacting with anti-£. coli recA 109 antibodies was determined by scanning densitometry of autoradiographies.

4.2.4 Labeling Plasmid DNA with Tritiated Nucleotide A single colony of £. coli harboring plasmid p-50 was inoculated into 5 ml of LB medium containing 50 ug/ml ampicillin and incubated at 37°C overnight with vigorous shaking. The 5 ml overnight culture was then inoculated into 100 ml LB medium containing 50 ug/ml ampicillin and cultured at 37°C with vigorous shaking until the culture reached OD^g = 0.4. After chloramphenicol was added to the culture at the concentration of 170 ug/ml, 10 mCi of tritiated dTTP (25 Curies/mmol) was added to the flask and the incubation continued for 12 hours in the shaking incubator. Following incubation, the cells were pelleted at 4,000 x g for 5 min, resuspended in 100 ml of 0.05 H Tris (pH 8.0) and 0.01 M EDTA, and pelleted as above. The pelleted cells were then resuspended in 2 ml of 25% sucrose and 0.01 H Tris (pH 8.0) followed by the addition of 0.3 ml of lysozyme solution (15 mg/ml in 0.25 M Tris, pH 8.0). After incubation for 5 min on ice, 0.6 ml of 0.25 M EDTA solution were added, and the solution was mixed by swirling the tube gently. The tube was then incubated on ice for 10 min and then 1.1 ml of Brij detergent solution containing 1% w/v Brij 58, o.4% Na deoxycholate, 0.0625 H EDTA, pH 8.0, 0.05 M Tris pH 8.0 was added. The tube was gently swirled followed by centrifugation 110 at 35,000 x g for 1 hour. The supernatant was removed and placed in a small beaker to which was added cesium chloride at 1 g/ml and ethidium bromide at 0.2 mg/ml. The cesium chloride solution was then centrifuged at 150,000 x g in a VTi 50 rotor for 24 hours. The plasmid DNA collected was then extracted three times with cesium chloride saturated butanol of equal volume to remove the ethidium bromide. The DNA was then placed in dialysis tube, dialyzed versus 1.5 liters of TE buffer (pH 8.0) for 16 hours at 4°C. Following the addition of 0.1 volume of 3 M sodium acetate (pH 5.4) and 2.5 volume of ethanol into the dialyzed DNA solution, DNA was precipitated at ~20°C overnight, and centrifuged at 12,000 xg at 4°C for 20 minutes. The DNA pellet was washed in 70% ethanol and re-pelleted as above. The precipitated DNA was dried in a speed vacuum and resuspended in 200 ul TE buffer. The activity of DNA was 0.5 X 104 cpm/ug DNA.

4.2.5 DNA Renaturation Assay DNA renaturation was carried out based on the SI assay for duplex DNA formation (Weinstock et al 1979). The £. coli recA reaction mixture contained: 20 mM Tris-HCl (pH 7.5) , 1 mM dithiothreitol, 10 mM MgClz, bovine serum albumin at 50 ug/ml, 10 mM KC1, 5% glycerol, heat denatured DNA (2 ug, 1 x 104 cpm) and 800 uM ATP. The reaction was performed with 1 ug of bacterial recA protein or 20 ug total protein isolated from Chiamvdomonas chloroplast stromal extracts at 37°C for 1 hour. Ill The reaction was stopped by the addition of 0.5 ml of SI nuclease buffer (300 mM NaCl, 50 mM sodium acetate, pH 4.6 and 1.0 mM zinc acetate) , 10 ug of heat denatured calf thymus DNA, and 400 units of SI nuclease. Incubation was continued at 37°C for 30 minutes. DNA was precipitated by adding 0.5 ml of 10% trichloroacetic acid and carrier calf thymus DNA was added to 40 ug/ml, and pelleted at 12,000 X g for 10 minutes at 4°C a in roicrofuge. The acid-precipitable radioactivity was measured by liquid scintillation counting. The above renaturation reaction was also performed in the absence of ATP to determine the effect of ATP on recA catalyzed ssDNA reannealing. In order to determine optimal reaction conditions for the chloroplast recA like protein, DNA renaturation assays were performed as described above except for the following variation: (1) reaction mixture pH from 5.5 to 6.5 [using 2- (N-) ethanesulfonic acid and 7-8.5 (suing tricine)? (2) reaction temperature from 20 to 40°C; (3) ATP concentration from 0 to 5.0 mM; (4) monovalent salt concentration, NaCl or KC1, at concentrations from 0 to 2 0 mM; and (5) divalent salt concentrations, CaCl2 or MgCl2, at concentrations from 0 to 20 mM.

4.2.6 Purification of RecA-like Protein Stromal proteins from chloroplasts of Chlamydomonas were extracted as described above. Proteins were precipitated with various concentrations of solid ammonium sulfate. After 1 hour 112 on ice, the suspension was centrifuged at 12,000 xg for 20 min. The supernatant was removed and the precipitated proteins were resuspended in P buffer containing 20 mM potassium phosphate (pH 6.5), 10% (v/v) glycerol, 1 mM dithiothreitol and 0.1 mM EDTA, followed by minidialysis against same P buffer at 4°C for 2 hours. Ammonium sulfate precipitated proteins (2.4 M) were further fractionated on either: 1) a cellulose phosphate column (0.8 X 10 cm) that was equilibrated with P buffer and eluted with step gradients of: 100, 200, 300, 400 and 500 mM KC1, or 2) a DEAE sephadex column (0.8 X 10 cm) pre-equilibrated with P buffer and proteins were eluted from the DEAE column with 0, 200 and 500 mM KC1.

4.2.7 Preparation of pxi3 s s d n a A single colony of J2. coli HB101 containing plasmid pX13 (6.2 kb) having an 3.0 kb Xbal fragment of psbA gene was used to prepare pX13 ssDNA. The single colony was inoculated into 5 ml of 2 X TY medium (1.6% tryptone, 1% yeast extract, 0.5% NaCl) and VCSM13 helper phage was added at 108 plock forming units/ml. The culture was grown at 37°C for 2 hours with vigorous shaking. Kanamycin was then added to 70 ug/ml and the culture was continuously grown for 17 hours. 1.5 ml of cultures transferred into a microfuge tube and centrifuged at 12,000 x g in a microfuge at 4°C for 5 minutes. 1.2 ml of supernatant was transferred to a tube and 200 ul of a solution of 20% polyethylene glycol (PEG 8,000) in 2.5 M NaCl was 113 added, followed by gentle vortexing and sitting on ice for 15 minutes. The bacteria phage particles were recovered by precipitation at 12,000 x g for 5 minutes at 4°C in a microfuge. The pellet was resuspended 400 ul 0.3 M sodium acetate, pH 6.0/1 mM EDTA by vigorous vortexing. After extraction with phenol and chloroform, the aqueous phase was transferred into a fresh microfuge tube, 1 ml ethanol was added. The DNA was precipitated by centrifugation at 12,000 x g at 4°C for 15 minutes. The pellet was resuspended in 70% ethanol and pelleted at 12,000 x g at 4°C for 10 minutes. After the pellet was dry in speed vacuum, the DNA was resuspended in 25 ul HzO.

4.2.8 Preparation of r32P1- dsDNA Bam HI digested pX13 DNA was dephosphorylated in the 100 ul mixture containing 0.2 M Tris-HCl (pH 8.0) , 10 mM MgCl2 and 10 mM ZnCl2, 1.0 unit of calf intestine alkaline phosphatase at 37°C for one hour. The reaction was stopped by adding 100 ul stop buffer containing 25 mM Tris-HCl pH 7.8, 10 mM EDTA, 0.2 M NaCl and 0.5% SDS followed by heating to 68°C for 30 minutes. 100 ul of 7.5 M ammonium acetate and 700 ul of ethanol were then added into the mixture and incubated at - 70°C for 3 hours. The dephosphorylated DNA was precipitated at 12,000 x g for 30 minutes at 4°C. The pellet was resuspended in 70% ethanol and centrifuged at 12,000 x g at 4°C for 10 min. The pellet was dried in speed vacuum and resuspended in 114

10 Ul H20. The 5'-hydroxyl ends of 25 pinoles of 5'-termini of pX13 were labeled in 25 ul mixture containing: 50 mM Tri-HCl (pH 7.6), 10 mM MgCl2, 5 mM DTT, 0.1 mM spermidine, 0.1 mM EDTA, 1 uM ATP, 100 uCi of r-[32P]-ATP and 20 units of T4 kinase at 37°C for 60 minutes. The reaction was stopped by adding 200 ul 2.75 M ammonium acetate, followed by adding 760 ul ethanol and stored at -70°C for 3 hours. The DNA was precipitated at 12,000 x g at 4°C for 15 min. The pellet was resuspended in 70% ethanol and pelleted as above. The pellet was dried in speed vacuum and resuspended in 3 0 ul HzO. The unincorporated r-(32P)-ATP was removed by running the DNA solution through the G-50 sephadex column which was preequilibrated with STE buffer containing 10 mM Tris.HCl (pH 8.0), 100 mM NaCl and 1 mM EDTA (pH 8.0).

4.2.9 Assay of Three-Stranded DNA The assay was carried out basically according to the method of Rao et al (1991). Presynaptic filaments were formed by incubating 200 ng circular ssDNA with 2 ug £. coli recA protein and 0.83 uM single-stranded DNA binding protein at 37°C for 12 minutes in a reaction mixture containing 3 3 mM Tris-HCl (pH 7.5), 12 mM Mgcl2, 2 mM dithiothreitol, 1.2 mM ATP, 8 mM phosphocreatine, creatine phosphokinase (10 units/ml) and bovine serum albumin (100 ug/ml, nuclease free). Paring and strand-exchange reactions were initiated by adding 115 0.5 ug linear 3zP-5'end-labeled duplex pX13 DNA. The reaction was carried out at 37°C for 25 min and stopped by adding EDTA to 20 mM, SDS to 0.5%, and proteinase K to 100 ug/ml, followed by incubation at 37°C for additional 20 min. The sample with blue dextran was filtered through a 10 x 0.8 cm CL-4B column preequilibrated with the buffer containing 10 mM Tris-HCl (pH 8.0), 0.6 M NaCl, 1 mM EDTA (pH 8.0) and 0.1% sarkosyl. DNA in the collected blue fraction was reprecipitated by ethanol as described before. The joint molecules were analyzed by electrophoresis on 1.0% agarose gel with TBE buffer containing 1.1% Tris-base, 0.55% boric acid and 0.01 M EDTA (pH 8.0), after which the gel was dried and exposed for autoradiography. To test strand exchange activities of the Ch1amydomonas recA- like protein the same strand exchange assay was performed except that 10 mM MgCl2 and 10 ug proteins from crude extract, ammonium sulfate fraction or DEAE fraction were used and the reaction temperature was 25°C.

4.3 RESULTS 4.3.1 Identification of "RecA-like" Protein in the Chloroolast of Chlamvdomonas Since recA protein is involved in recombination and DNA repair in bacteria, it is also interesting to know whether "recA" is also present in chloroplasts. Chlamvdomonas strain CC-406 was chosen as the experimental organism, since it is cell wall-less and is a good source of intact chloroplasts. 1 16 The stromal proteins were extracted from purified chloroplasts of cw-15 cells and separated by SDS gel electrophoresis. The stromal extracts were then western blotted onto nitrocellulose followed by immunodetection with antisera against £. coli recA protein. Only one band of 27 kd was detected as shown in Fig 11. This result indicates that the £. coli recA protein and a chloroplast protein are immunochemically related. It has reported that £. coli recA protein has a molecular weight of 37842 (Horii et al., 1980) which is larger than this chloroplast protein, indicating the difference between them. It has previously been shown that DNA gyrase inhibitors will induce the synthesis of recA protein in £. coli. In order to determine whether such treatments would have a similar effect on the Chlamvdomonas 27 kd protein, the level of the "recA" immuno-reactive protein was quantified by western blot analyses of chloroplast extracts from cells treated with novobiocin for various time intervals. As shown in Fig 12, the level of cross-reacting protein in novobiocin treated cells increased two fold compared with non-treated cells over a period of three hours and was maintained at these levels in the presence of novobiocin. This result is similar to that found in £. coli cells treated with nalidixic acid (Thoms and Wackernagel, 1987) , suggesting that this recA-like protein may be induced in chloroplasts by treatments which turn on the SOS response system in £. coli. 12 3 4 5 6 97 kd -

66 kd -

45 kd - 34 kd -

- 27 KD 22 kd -

Figure 11. Western blot analysis of stromal polypeptides isolated from Chlamvdomonas cells probed with the E. coli recA protein antibody. Lane 1 and lane 2, one day culture; lane 3, and lane 4, two day culture; lane

5 and lane 6, three days culture. Samples were loaded on the basis of equal amounts of 50 ug protein. 117 Figure 12. Effects of novobiocin on the levels of recA-like protein from the chloroplast of Chlamvdomonas detected with E. coli recA antibody (1:1,000 dilution). Stromal proteins (50 ug) were applied to nitrocellulose and immunodetected with antisera against £. coli recA protein. The abundance of the protein was determined by scanning densitometry of the autoradiographies. Data are represented as the means of three measurements.

118 Maximum RecA Level 105 95 85 65 75 55 45 0 ■*— -Novobiocin Hours Treatment Treatment Hours rigur* 12. rigur* 3 -- a ** ■•‘Novobiocin 6 9 12 120 4.3.2 Renaturation of DNA One of the processes catalyzed by recA protein is the renaturatlon of denatured, double-stranded DNA (Weinstock et al., 1979). RecA protein Isolated from bacteria or protein from stromal extracts of Chiamvdomonas cw-15 cells was incubated with 2 ug of tritium labeled heat-denatured plasmid p50. The extent of duplex DNA formed during the renaturation reaction was determined by conversion of the single-stranded p50 DNA to SI nuclease resistant double-stranded p50 DNA. As shown in Fig. 13, incubation of denatured DNA with recA protein or chloroplast protein extracts from Chlamvdomonas cells produced SI nuclease-resistant double-stranded DNA. Incubation of denatured DNA in the absence of recA protein or chloroplast protein did not result in formation of SI nuclease-resistant double-stranded DNA. These result suggests that there is a recA-like enzyme activity in Chlamvdomonas chloroplasts, which catalyzes DNA renaturation of single­ stranded DNA. The results also demonstrate that the renaturation reaction was stimulated by ATP although there was some duplex DNA formed in the presence of chloroplast protein without ATP (Fig. 13) . Similarly, a 2 fold increase formation of DNA aggregates stimulated by ATP was observed in £. coli recA catalyzed DNA renaturation (Weinstodk et al., 1979). Therefore, the formation of Sl-nuclease-resistant dsDNA catalyzed in vitro by a chloroplast stromal protein strongly Figure 13. Annealing of single-stranded DNA by a protein in stromal extracts from Chiamvdomonas chloroplasts. RecA protein isolated from bacteria or protein from stromal extracts of Chlamvdomonas reinhardtii was incubated with 2 ug of tritium labeled, heat-denatured plasmid p50 and the amount of SI nuclease-resistant acid precipitable dsDNA was determined by liquid scintillation counting. R-A: bacterial recA without added ATP; R+A: bacterial recA plus 800 uM ATP; C-A: Chlamydomonas protein without ATP; C+A: Chlamvdomonas protein plus 800 uM ATP.

121 Maximum Level 100 110 20 60 90 30 40 80 70 50 10 0 GZZZD r - a 0:00 2 3 E2Z r + Flgur* 13 Flgur* a ie (hour) Time 3 S E c - a 1:00 A O 123 suggests that Chlamydomonas chloroplasts may contain a recA- like protein functional in recombination.

4.3.3 Determination of Optimal Conditions for Chloroplast Protein Catalyzed DNA Renaturation Reaction In order to determine the optimal conditions for the renaturation of ssDNA by chloroplast extracts, we varied pH, temperature, ATP, monovalent and divalent salt concentrations. These assays were performed under conditions which would be saturating for the quantity of chloroplast protein used. The first factor assayed was pH. The renaturation reaction was carried out at pH 5.5, 6.0, 6.5, 7.0, 7.5, 8.0 and 8.5. As shown in the Fig. 14, the optimal pH was 7.5. This is unlike single-stranded DNA-stimulated ATP hydrolysis in £. coli. which occurs over a wide range of pH values (Weinstock et al., 1981; McEntee et al., 1981). However, the optimal pH (7.5) for the Chlamvdomonas protein catalyzed reaction was similar to the chloroplast stromal pH. The effect of temperature on the renaturation activity is shown in Fig. 15. The optimal temperature was 25°C. There was a 60% or 50% decrease in activity at 2 0 or 30°C respectively. The temperature for recA activity corresponds to the optimal temperature for Chlamvdomonas growth. Similarly, in bacteria, the optimal temperature for recA activity (37°C) is identical to that optimal for normal growth. Figure 14. Effect of pH on Chlamvdomonas recA-like protein catalyzed DNA renaturation. The reaction was performed in 2 0 mM MES buffer at various pH (5.5, 6.0, 6.5) or in 20 mM Tricine buffer at pH 7.0, 7.5, or 8.0. The reaction mixture also contained: 1 mM dithiothreitol, 10 mM MgCl2, bovine serum albumin at 20 ug/ml, 10 mM KC1, 5% glycerol, heat denatured DNA (2 ug, 4 X 104 cpm) , 800 uM ATP, 50 ug total protein isolated from Chlamydomonas. The highest specific activity (cpm/ug protein) at pH 7.5 was used as maximal level of SI nuclease resistant double-stranded DNA. Data are represented as means of three experiments. The error bar is a standard deviations of three replications.

124 of maximal SI nuclease resistant dsDNA 100 110 20 40 90 30 50 60 80 70 10 0

* ------sMI 5.5 “T*1 __ Figure Figure 4 1 . Figure 15. Effect of temperature on Chlamvdomonas recA-like protein catalyzed DNA renaturation. The reaction was performed in Tris-HCl buffer (pH 7.5) at various temperatures (20, 25, 35, and 40°C) . The reaction mixture also contained: 1 mM dithiothreitol, 10 mM MgCl2, bovine serum albumin at 50 ug/ml, 10 mM KCl, 5% glycerol, heat denatured DNA (2 ug, 4 X 104 cpm), 800 uM ATP, 2 0 ug total protein isolated from Chlamydomonas. The highest specific activity (cpm/ug protein) at 25°C was used as maximal level of SI nuclease resistant double-strand DNA. Data are represented as means of three experiments. The error bar is a standard deviation of three replications.

126 of maximal SI nuclease reaistant dsD 100 110 120 80 90 60 30 70 40 50 20 10 ------20 25 eprtr *C Temperature Figure 15. Figure 30 35 40 128 The optimal ATP concentration for recA activity was also determined. As shown in Fig. 16, the maximum activity of the recA-like enzyme was obtained at 1.5 mM ATP. The enzyme had classical Michaelis-Menten kinetics. The measured km and Vmax for ATP were 0.83 mM and 62.5 ug dsDNA/mg protein/hour respectively (Fig. 17). The effects of salts on DNA renaturation catalyzed by recA-like protein was also determined. As shown in Fig. 18, the activity of chloroplast recA-like enzyme increased 5 fold by the addition of 5 mM NaCl compared to no salt in the reaction mixture. A similar effect was also observed with KCl. At a concentration of 10 mM KCl a 7-fold increase of recA-like protein activity was observed in contrast to the reaction carried out in the absence of salt (Fig 19) . Addition of either Na+ or K* above the optimum concentration resulted in decreased activity (Fig 18, 19) . Optimal recA activity was observed at a MgCl2 concentration of 10 mM (Fig. 20). However, at 20 mM of Mg2*, the specific activity dropped to 30 percent of the maximal level of the activity. Similarly, the concentration of CaCl2 required for optimal recA activity was also 10 mM. The specific activity declined to 20 per cent of maximal activity when the concentration of Ca2* reached to 20 mM (Fig. 21) . Figure 16. Effect of ATP concentration on Chlamvdomonas recA- like protein catalyzed DNA renaturation. The reaction was performed in Tris-HCl buffer (pH 7.5) at various concentration

O f ATP (0, 0.1, 0.5, 1.0, 1.5, 2.0, 3.0, 4.0, and 5.0 mM). The reaction mixture also contained: 1 mM dithiothreitol, 10 mM MgCl2, bovine serum albumin at 20 ug/ml, 10 mM KCl, 5% glycerol, heat denatured DNA (2 ug, 4 X 10* cpm) , 50 ug total protein isolated from Chlamvdomonas. Data are represented as means of three experiments.

129 5000

4000

3000 a a o 2000

1000

2 2.5 3 3.5 ATP (mM)

Flyur* 1«. 131

1/V

-1/Km

.2 0 1 2 1/Catp)

Figure 17. Michaelis-Menten kinetics of ATP dependent renaturation reaction catalyzed by recA-like protein. 1/V (ug dsDNA x 10*2/mg protein/hour) is plotted as a function of

1/[ATP], the intercept on the y-axis is 1/Vmax, and the intercept on the x-axis is -1/km. Figure 18. Effect of NaCl concentration on Ch1amvdomonas recA- like protein catalyzed DNA renaturation. The reaction was performed in Tris-HCl buffer (pH 7.5) at various concentrations of NaCl (0, 1, 2, 5, 10, 15, 20 mM) . The reaction mixture also contained: 1 mM dithiothreitol, bovine serum albumin at 2 0 ug/ml, 5% glycerol, heat denatured DNA (2 ug, 4 X 104 cpm) , 1.5 mM ATP, 50 ug total protein isolated from Chlamydomonas. The highest specific activity (cpm/ug protein) at 10 mM was used as maximal level of SI nuclease resistant double-stranded DNA. Data are represented as means of three experiments. The error bar is a standard deviation of three replications.

132 1 5 0 5 20 15 10 5 2 1 0 # of maximal SI nuclease resistant dsDNA 100 110 30 60 80 40 50 20 90 70 10 0

m

Figure l®. Figure

) M m ( l C a N 133 Figure 19. Effect of KCl concentration on Chlamvdomonas recA- like protein catalyzed DNA renaturation. The reaction was performed in Tris-HCl buffer (pH 7.5) at various concentrations of KCl (0, 1, 2, 5, 10, 15, 20 mM) . The reaction mixture also contained: 1 mM dithiothreitol, bovine serum albumin at 2 0 ug/ml, 5% glycerol, heat denatured DNA (2 ug, 4 X 104 cpm) , 1.5 mM ATP, 50 ug total protein isolated from Chlamvdomonas. The highest specific activity (cpm/ug protein) at 10 mM was used as maximal level of SI nuclease resistant double-stranded DNA. Data are represented as means of three experiments. The error bar is a standard deviation of three replications.

134 < % Q«i 110 *o 100 90 80

9 70 «k m *4 60 o o a 50 40 30 j M 20 « a 10 o 0 2 5 10 K C l ( m M ) Figure 20. Effect of MgCl2 concentration on Chlamvdomonas recA-like protein catalyzed DNA renaturation. The reaction was performed in Tris-HCl buffer (pH 7.5) at various concentrations of MgCl2 (0, 1, 2, 5, 10, 15, 20 mM) . The reaction mixture also contained: 1 mM dithiothreitol, bovine serum albumin at 50 ug/ml, 5% glycerol, heat denatured DNA (2 ug, 4 X 104 cpm) , 1.5 mM ATP, 20 ug total protein isolated from Chlamvdomonas. The highest specific activity (cpm/ug protein) at 10 mM was used as maximal level of si nuclease resistant double-stranded DNA. Data are represented as means of three experiments. The error bar is a standard deviation of three replications.

136 of maxiaml SI nuclease resistant dsDNA 100 110 30 80 90 40 60 20 50 70 10 10 0

1 5 1 1 20 15 10 8 5 2 1 0 Figure 20. Figure l (mM) M m ( Cl, g M Figure 21. Effect of CaCl2 concentration on Chlamydomonas recA-like protein catalyzed DNA renaturation. The reaction was performed in Tris-HCl buffer (pH 7.5) at various concentrations of CaCl2 (0, 1, 2, 5, 10, 15, 20 mM) . The reaction mixture also contained: 1 mM dithiothreitol, bovine serum albumin at 50 ug/ml, 5% glycerol, heat denatured DNA (2 ug, 4 X 104 cpm), 1.5 mM ATP, 20 ug total protein isolated from Chlamydomonas. The highest specific activity (cpm/ug protein) at 10 mM was used as maximal level of SI nuclease resistant double-stranded DNA. Data are represented as means of three experiments. The error bar is a standard deviation of three replications.

138 z< o 110 100 a 90 «* « 80 70 u« 60 O 9 a 50 co 40

* 5 8 10 CaCl (mM)

rlgur* 21. 140 .4*3*4 Partial Purification of RecA-like Protein The recA-like protein was partially purified from crude chloroplast stromal extracts. The first method used was ammonium sulfate fractionation. The proteins in the crude extracts were precipitated at different concentrations of ammonium sulfate. After dialysis of each fraction, the specific activity in each fraction was determined using the ssDNA renaturation assay. As shown in Table 9, protein in the 2.4 M ammonium sulfate fraction had the highest specific activity. Since there was little activity in fractions precipitated at less than 1.2 M ammonia sulfate, we first removed contaminating proteins from the crude extract by precipitating with 1.2 M ammonia sulfate. RecA activity was then precipitated with 2.4 M ammonia sulfate. The next step in the purification of the recA protein was to use ion-exchange chromatography (including cellulose phosphate and DEAE sephadex columns). The proteins in 2.4 M ammonium sulfate fractions were loaded on the ion-exchange column and eluted from the column with KC1 at various concentrations. The results showed that the protein appears to bind to cellulose phosphate and DEAE sephadex columns and that each fraction has measurable recA activity (Table 10, 11). As shown in Fig. 20, the results demonstrated that ammonium sulfate fractionation followed by DEAE-sephadex purification gave a 10-fold increase in specific activity (Fig 22). The some amount of proteins in crude extracts, ammonium sulfate fraction and DEAE fractions 141

Table 9. Ammonium sulfate fractionation of Ch 1 amvdomonas recA- like protein. Stromal proteins were precipitated with various concentrations of ammonium sulfate, followed by minidialysis against P buffer containing 20 mM potassium phosphate (pH 6.5), 10% (v/v) glycerol, 1 mM dithiothreitol and 0.1 mM EDTA at 4°C for two hours.

0 202 57.4 3.5 150 357 75.6 4.7 450 345 71.4 4.8 600 285 75.6 3.8 1200 356 88.2 4 . 0 2400 1243 74.2 16.8 142 Table 10. Cellulose phosphate purification of recA-like protein. Proteins precipitated in 2.4 M ammonium sulfate fraction were fractionated on a cellulose phosphate column (0.8 x 10 cm) that had been equilibrated with P buffer containing 20 mM potassium phosphate (pH 6.5), 10% (v/v) glycerol, 1 mM dithiothreitol and 0.1 mM EDTA. The proteins eluted with step gradients of 0, 100, 200, 300, 400 and 500 mM KCl were precipitated by 2.4 M ammonium sulfate, followed by minidialysis against P buffer at 4°C for two hours.

KCl Fraction Precipitable Total protein Specific activity (mM) dsDNA (cpm) (ug) (cpm/ug)

0 2060 128 16 100 17761 230 77 200 10159 218 46 300 5758 163 35 400 8709 157 55 500 4715 131 36 Table 11. DEAE sephadex purification of Chlamvdomonas recA-like protein. Proteins precipitated in 2.4 M ammonium sulfate fractions were loaded on a DEAE sephadex column (0.8 x 10 cm) that had been equilibrated with P buffer containing 20 mM potassium phosphate (pH 6.5), 10% (v/v) glycerol, 1 mM dithiothreitol and 0.1 mM EDTA. The proteins were eluted with step gradients of 0, 2 00, and 500 mM KCl and were precipitated by 2.4 M ammonium sulfate followed by minidialysis against P buffer at 4°C for two hours. Activity was determined by renaturation assay.

KCl fraction Precipitable Protein Specific activity

(mM) dsDNA (cpm) (ug) (cpm/ug)

eluate 2330 102 23

200 8184 106 77

500 5222 122 43 143 Figure 22. Annealing of single strand DNA by Chiamvdomonas recA-like protein (20 ug) in crude extract, ammonium sulfate fraction and DEAE fraction. RecA-like protein purified from each step was incubated with 2 ug of tritium labeled heat- denatured plasmid p50 and the amount of SI nuclease-resistant precipitable DNA was determined by liquid scintillation counting. Data are represented as means of three experiments. The error bar is a standard deviation of three replications.

144 ct Mltfifi

o mxml I ulae eitn dsDNA resistant nuclease SI maximal of % 120 146 were then separated by SDS-PAGE. As shown in Fig. 23, only one polypeptide(2.7 kd) was enriched in the DEAE fraction. The proteins separated on the SDS-PAGE gel were also western blotted and iumunodetected with £. coli recA antibodies, followed by densitometry analysis. We observed a 10 fold enrichment of the recA, immunodetecable 27 kd protein in the DEAE fraction relative to crude extracts (Fig. 24). These results suggest that the 27 kd protein, recognized by £. coli recA antibodies, catalyzes the renaturation of ssDNA (Fig.

22) .

4.3.5 Formation of Three-Stranded DNA £. coli protein catalyzed strand exchange between circular pX13 ssDNA and homologous linear duplex pX13 was carried out according to the procedure of Rao et al. (1991). As shown in Fig. 25, two bands were observed in the presence of £. coli recA while only one band was shown in the absence of the recA. The distinct band with higher molecular weight indicated that a 3 stranded molecule was formed by ssDNA and dsDNA. The formation of this joint molecule was catalyzed by the £. coli recA. Using chloroplast proteins from the crude stromal extract, the ammonium sulfate fraction and the DEAE fraction, the same strand exchange assay was performed. As shown in Fig. 26, there were no extra bands formed, indicating there were no 3 stranded molecules formed. Since £. coli recA can catalyze the formation of joint molecules, this result may Figure 23. Protein separation by sodium deodecyl sulfate (SDS)-polyacrylamide gel electrophoresis. Identical amounts (100 ug) of crude extract (1), the 2.4 M ammonium sulfate fraction (2) and the DEAE fraction (3) were separated electrophoretically at 10 mA for 15 hours at 4°C and stained with Coomossie Blue.

147 Figure Figure 24. Purification of recA-like protein from Chlamvdomonas chloroplasts. Following ammonium sulfate precipitation and DEAE purification equal amounts of stromal protein were separated by SDS gel electrophoresis and western transferred onto nitrocellulose. The recA-like proteins were immunodetected with E^. coli recA antibodies and quantified by densitometry. Only a single immunoreacting band at 27 kd was detected.

149 of Reletive Desity 110 HH HH rd (nh Crude Figure 24.Figure 4 )2S 0 4, M W

M B D 150 Figure 25. Formation of joint molecules by £. coli recA. Presynaptic filaments were formed by incubating 200 ug circular ssDNA with 2 ug £. coli recA protein and 0.83 uM single-stranded DNA binding protein at 37°C for 12 minutes in reaction mixture containing 33 mM Tris-HCl (pH 7.5), 12 mM MgCl2, 2 mM dithiothreitol, 1.2 mM ATP, 8 mM phosphocreatine, creatine phosphokinase (10 units/ml) and bovine serum albumin (100 ug/ml, nuclease free). Paring and strand-exchange reactions were initiated by adding 0.5 ug linear 32P-5'-ends- labeled duplex pX13 DNA. The reaction was carried out at 37°C for 25 min and stopped by adding EDTA to 2 0 mM, SDS to 0.5%. After deproteinization with proteinase K and filtration through CL-4B column. The sample were analyzed by electrophoresis on 1.0% agarose gel with TBE buffer (A). The gel was then dried and exposed for autoradiography (B) . M, marker; 1, the reaction performed in the presence of £. coli recA; 2, the reaction performed in the absence of £. coli recA.

151 152

M 1 2

23.1 Kb 9.4 kb 6-6 Kb

1 2

Figure 25. Figure 26. Formation of joint molecules by recA-like protein. The assay was performed by incubating 200 ug circular ssDNA with 10 ug DEAE fraction protein and 0.83 uM single-stranded DNA binding protein at 37°C for 12 minutes in reaction mixture containing 33 mM Tris-HCl (pH 7.5), 10 mM MgCl2, 2 mM dithiothreitol, 1.2 mM ATP, 8 mM phosphocreatine, creatine phosphokinase (10 units/ml) and bovine serum albumin (100 ug/ml, nuclease free). Paring and strand-exchange reactions were initiated by adding 0.5 ug linear 32P-5•-ends-labeled duplex pX13 DNA. The reaction was carried out at 25°C for 25 min and stopped by adding EDTA to 20 mM, SDS to 0.5%. After deproteinization with proteinase K and filtration through Cl- 4B column. The sample were analyzed by electrophoresis on 1.0% agarose gel with TBE buffer (A). The gel was then dried and exposed for autoradiography (B) . M, marker; lanel, the reaction performed in the absence of recA-like protein; lane 2, without ATP in the reaction mixture; lane 3, crude extracts; lane 4, ammonium sulfate fraction; lane 5, DEAE fraction.

153 154 A

M 1 2 3 4 5

23.1 9.4 6.6

2.2 2.0

B 1 2 3 4 5

Figure 26. 155 suggest that the strand exchange reaction in Chlamvdomonas chloroplasts is more complex than in £. coli.

4,4 DISCUSSION Our objective has been to identify and characterize the proteins which may be involved in chloroplast DNA recombination and repair in chloroplasts. The stromal proteins from Ch1amvdomona s chloroplasts were isolated, partially purified and characterized for this purpose. A recA-like protein has been found based on the following evidence. First, a single polypeptide was found to cross react with antibodies raised against the £. coli recA protein although its size (27 kd) is smaller than E. coli recA (38 kd) (Horii et al. 1980). Since recA protein is involved in recombination and DNA repair in £. coli (Thoms and Wackernagel, 1987), our result suggests that the immunochemically related protein in Chlamydomonas may be also involved in DNA recombination. Second, the steady state level of the chloroplast protein which cross-reacts with £. coli recA antibodies increased following novobiocin treatment. This result is similar to that for recA in £. coli following treatment with DNA gyrase inhibitor (Thoms and Wackernagel 1987). Third, it has been shown that chloroplast stromal extracts contain proteins which accelerate the renaturation of ssDNA, as does £. coli recA protein. Finally, our results showed that this DNA renaturation reaction was enhanced in the presence of ATP as is the case for the £. coli 156 recA enzyme. The recA-like protein was partially purified, and the immunologically cross-reacting protein (27 kd) was enriched 10 folds in the most active fraction. A ten fold purification was obtained, as indicated by the increase in specific activity observed with the DNA renaturation assay. Interestingly, the polypeptide cross reacting with the £. coli recA antibodies was also enriched 10 fold during purification. Some of the methods used for the purification of this recA-like protein are similar to those used to purify £. coli recA (Weinstock, 1979). However, the chloroplast protein has not been purified to homogeneity. It should also be noted that recA-like protein in Chlamvdomonas chloroplasts has molecular weight of about 27 kd, while £. coli recA protein has a molecular weight of 38 kd (Horii et al. 1980). This difference in size may be due to evolutionary diversity. In £. coli. the initial event in the SOS response is the activation of recA by DNA damage. Interestingly, our results have shown that the level of recA-like protein in chloroplasts can be increased by treatment with the DNA gyrase inhibitor novobiocin. Under normal growth conditions, recA protein in £. coli is expressed at a basal level. Upon induction of the SOS system, this level is increased 20 fold (Salles and Paoletti, 1983; Salles et al., 1983). The activated recA protein interacts with the repressor protein coded by the lexA gene. This interaction then triggers proteolytic cleavage of the 157 repressor to allow more SOS genes to be expressed for effective DNA repair. Although not much is known about the relationship between the novobiocin treatment and the change in the level of chloroplast recA-like protein, the induction event may play an important role in the chloroplast for DNA repair. It is noted, however, that not all recA expression systems are regulated by lexA (Cox and Lehman 1987). Another interesting feature of recA-like protein found in chloroplasts of Chiamvdomonas is associated with renaturation of complementary single-stranded DNA. In addition, this renaturation reaction, presumably catalyzed by the recA-like protein is stimulated by ATP. These observations suggest that recA-like protein in chloroplasts has characteristics similar to those of E. coli recA protein (Weinstock et al., 1979; Bryant and Lehman, 1985; McEatee, 1985). It has been reported that ATP causes a two fold increase in the stoichiometry of binding of recA protein to ssDNA, implying that a recA monomer contains two DNA-binding sites (Bryant et al., 1985). Since two DNA-binding sites make it possible for recA protein to bind two single stranded DNAs at the same time the possible mechanism of DNA renaturation catalyzed by chloroplast recA- like protein could be characterized by the intermediate complex between a recA-ssDNA and a second ssDNA molecule. By determining the optimal conditions for the DNA renaturation assay, the effects of various factors on recA activity have been tested. Our results showed that the optimal 158 conditions of pH, concentration of monovalent salt, and divalent salt are close to the stromal environment in vivo. The optimal temperature (25°C) for the renaturation reaction is the same as that favored by the alga for growth, whereas bacteria grow optimally at 37°C. However, the optimal ATP concentration (2 mM) is similar to that required for £. coli recA activities. All these results indicate that the recA-like protein in Chlamvdomonas chloroplasts acts like an enzyme. The recA-like activity is not unique to JIU. coli. Proteins with properties similar to recA protein of JjL*. coli appear to be widely distributed among bacteria (Eitner et al., 1982). RecA analogues have been found in photosynthetic cyanobacteria which play a role in homologous recombination and UV resistance (Geogghegan and Houghton, 1987; Murghy et al., 1987; Owttrim and Coleman, 1987). In the lower eukaryote Ustilaao mavdis, a recA-like protein has been purified based on its ability to reanneal complementary single strands of DNA (Kmiec and Holloman, 1982). This protein promotes the uptake of single stranded DNA by duplex DNA in a homology dependent reaction in the presence of ATP (Kmiec and Holloman, 1984). RecA-like activities have been detected in mammalian cells as well. Several groups have reported a strand transfer activity in extracts of fibroblasts from patients (Kenne and Lindquist, 1984), extracts of human B lymphoblast (Hsieh et al., 1986) and extracts of Hella cells (Cassuto et al., 1987). In extracts of rat FR 3T3 cells, a protein has also been found to 159 be Immunological ly cross-reactive with the £. coli recA protein (Angulo et al., 1989). Unfortunately, there is no report about recA-like protein in plants. However, in the green alga Chlamvdomonas. treatment with the thymidine synthesis inhibitor fluorodeoxyuridine results in specific reduction of chloroplast DNA content and an increase the level of transmission of chloroplast DNA for mating type (mt'J in the crosses with mt*. This enhancement of recombination events is similar to SOS response observed in prokaryotic organisms when DNA synthesis is inhibited. Our studies described in Chapter III have shown that reduction of the chloroplast genome copy number in Chlamvdomonas increases the frequency of chloroplast transformation. However, it is not clear if this is due to lowered chloroplast DNA copy number or induction of a chloroplast DNA recombination system in Chlamvdomonas chloroplasts. Recently, the evidence supporting the presence of a recombination system in chloroplasts has come from the identification of chloroplast open reading frames homologous to the £. coli recA and uvr genes which are involved in recombination and DNA repair. Our studies have showed further evidence to suggest that an ATP dependent recA-like activity in Chlamvdomonas chloroplast stromal extracts may be involved in recombination process. Using the strand exchange method of Rao et al. (1991) the recombinase activity of £. coli recA and chloroplast recA-like 160 protein was also determined. The purified recA protein from £. coli has been demonstrated its ability to promote homologous pairing and formation of 3 stranded DNA molecules from ssDNA and homologous dsDNA. However, the joint molecules were not formed when Chiamydomonas stromal proteins were used. This result may imply that DNA recombination mechanisms in Chlamydomonas chloroplasts is different from that in £. coli or requires different strand exchange conditions 4a vitro. The DNA homologies and protein similarities between bacterial recA and Chlamvdomonas chloroplast recA-like protein does not prove that a functionally similar recA gene is also present in chloroplasts of higher plants. Since the chloroplast genomes of algae are typically much larger than those of higher plants, they may have additional coding sequences. On the other hand, genes involved in recombination and DNA repair in higher plant chloroplast may have a sequence that is totally different from genes in bacteria and even in algae. However, the identification and characterization of a recA-like gene and its product from Chlamvdomonas can provide us with valuable insight into the biochemical machinery of some of the most fundamental of DNA transactions as well as the evolutionary history of chloroplast DNA. CHAPTER ▼ SUMMARY

The original goal of this study was to develop a stable chloroplast DNA transformation system. It was proposed that by lowering the chloroplast genomic copy number we could enhance the frequency of recombination or integration of DNA into the chloroplast genome by stimulus of the recombination and DNA repair systems and by facilitating the segregation of transformed genomes. In addition, it was our objective to characterize the mechanism of genetic recombination in chloroplasts.

5.1 Reduction c p DNA Content bv Treatment with c p DNA Synthesis Inhibitors To determine whether cpDNA steady-state levels could be effectively reduced in higher plants without causing cell death, we screened a number of potential cpDNA synthesis inhibitors in suspension cell cultures of Solanum nigrum for their effects on chloroplast and nuclear DNA content and cytotoxicity. These studies were carried out with cultures which were either grown to stationary phase or rapidly transferred so as to maintain them in an active state of cell

161 division and cpDNA replication. One of the effects of the inhibitors was a reduction in cell growth and viability. Analyses of the chloroplast and nuclear DNA content per gram fresh weight by dot blot hybridizations indicated that the reduction of cpDNA content was greatest at inhibitor concentrations which reduced cell growth by more than 50% but this depended on the culture conditions. For example, the two DNA gyrase inhibitors, nalidixic acid and novobiocin, were more effective in lowering cpDNA content in cultures which were transferred (2 X 4 days) once during the eight day incubation. Because several inhibitors were toxic to cell growth, the DNA content of treated cells was also determined on the basis of cell (protoplasts) number. Analyses of nuclear and cpDNA content per cell for each treatment indicated that only the DNA gyrase inhibitors, nalidixic acid, and novobiocin reduced cpDNA content. Neither inhibitor reduced nuclear DNA content. These results suggest that DNA gyrases participate in cpDNA replication. The selective reduction of cpDNA content in regenerable cultures may facilitate the generation and selection of cpDNA mutants or transformants from higher plants.

162 163 5. 2 Transformation fif S. nigrum plastids and Mutagenic

Effects of c p DNA Synthesis Inhibitors The ideal chloroplast transformation system would be based on random DNA integration into chloroplast genome or by homologous recombination. We used a psbA gene conferring atrazine resistance from Amaranthus hvbridus as a selectable marker for transformation purpose. Following reduction of cpDNA copy number with novobiocin, calli and leaves of wild type S. nigrum were bombarded with the atrazine resistant psbA gene fragment by high-velocity microprojectiles. After regeneration of plants under atrazine resistance selection, potential transformants were screened by restriction fragment length polymorphism (the psbA atrazine resistance gene has a unique restriction endonuclease site). Unfortunately, there was no positive evidence for chloroplast transformation. However, it was found that cpDNAs isolated from selected atrazine resistant transformants have a higher affinity for an oligonucleotide, which has an altered nucleotide responsible for atrazine resistance, than cpDNA from wild type. This result suggests that cpDNA synthesis inhibitors may have mutagenic effects on the cpDNA genome. To test this hypothesis, atrazine resistant plants regenerated from novobiocin treated and nontreated leaves of wild type S. nigrum were selected. The results have shown that novobiocin treatment increased frequency of mutagenesis up to 20 fold, suggesting that these treatments may be mutagenic. 164

5.3 Regulation of c p DNA Content and Transformation Frequency in Chlamvdomonas Based on our studies, the two DNA gyrase inhibitors, nalidixic acid and novobiocin, were shown to be effective in lowering copy number of chloroplast genome in £. nigrum with little effect on the nuclear DNA content. Bacterial cells treated with nalidixic acid have higher DNA recombination and mutation frequencies (Thomas and Wadkernagel, 1987). These observations suggest that nalidixic acid and novobiocin treatment of plant cells may also affect cpDNA recombination processes. In support of this hypotheses, it is noted that treatment of Chlamvdomonas with 0.5 mM Fudr increased the frequency of cpDNA transformation (by homologous recombination) while also selectively lowering the cpDNA content (Boynton et al. 1988). In this study, chloroplast psbA gene deletion mutants of Ch1amvdomonas were treated with Fudr as well as novobiocin to reduce the number of chloroplast genomes. For the determination of nuclear and cpDNA content, total DNA extracted from the same number of cells was blotted onto nitrocellulose and hybridized with a chloroplast encoded gene (psbA) as well as the small subunit of Rubisco gene (nuclear encoded). Signals were quantified by scanning densitometry. The results demonstrated that 100 uM novobiocin selectively reduced cpDNA content in Ch1amvdomonas without reduction in cell growth and viability. 165 Using the particle bombardment method the pretreated photosynthetically defective mutants were transformed with cloned wild type chloroplast psbA gene and restored their photosynthetic capacity. It was observed that pregrowth of the deletion mutants in Fudr or novobiocin increased transformation frequency 35 and 24 fold respectively. These results suggest that transformation of plant cpDNA by recombination mechanisms may be facilitated by treatment of cells with DNA gyrase inhibitor.

5.4 RecA-like Protein in Chlamvdomonas Chloroplasts Recently the evidence supporting the presence of a recombination system in chloroplasts has come from the identification of chloroplast which is homologous to the £. coli recA and uvr genes (Oppermann et al., 1989). Since recA is involved in recombination and DNA repair and recA protein synthesis can be promoted by nalidixic acid treatment of £. coli (Thomas and Wackernagel, 1987), we investigated whether recA protein(s) were also present in chloroplasts. Proteins were extracted from purified chloroplasts of Chlamvdomonas separated by SDS-PAGE and transferred onto nitrocellulose followed by immunodetection with antisera against £. coli recA. Only one band, a 27 kd protein, was detected. This result indicated that the £. coli recA protein and a chloroplast protein were immunochemically related. 166 It has previously been shown that DNA gyrase Inhibitors will induce the synthesis of recA in £. coli. in order to determine whether such treatments would have a similar effect on Chlamvdomonas the cells were treated with novobiocin and the level of the "recA" immuno-reactive protein was quantified at various time intervals after addition of novobiocin. It was demonstrated that the level of recA like protein in novobiocin treated cells increased to levels two fold higher than non­ treated cells over a period of three hours and was maintained at these levels in the presence of novobiocin. This result is similar to that found in E. coli cells treated with nalidixic acid (Thomas and Wackernagel, 1987), suggesting a role of this recA like protein in DNA repair mechanisms. One of the processes catalyzed by recA protein is the renaturation of denatured, dsDNA (Weinstodk et al., 1979). The extent of duplex DNA formed during the renaturation reaction was determined by conversion of the single-stranded p50 ((psbA) DNA to SI nuclease resistant dsDNA. Incubation of denatured DNA with recA protein or chloroplast protein extracts from Chlamvdomonas cells produced SI nuclease- resistant dsDNA. The results also showed that the renaturation reaction was stimulated by ATP. Similar to £. coli recA some duplex material is formed in the presence of chloroplast protein without ATP. The formation of SI nuclease-resistant DNA catalyzed in vitro by the recA like protein strongly suggests that Chlamvdomonas chloroplasts may have a recA-like 167 protein functional in recombination. To purify the recA-like protein from the crude chloroplast extracts the first method we used was ammonium sulfate fractionation. The renaturation assay demonstrated that proteins precipitated by 2.4 M (NH4)2S04 had the highest specific activity. The protein was further purified by ion exchange chromatography. As shown by activity assays a 10 fold increase in specific activity was achieved. The most active fractions contained only several polypeptides. One of them cross reacts with £. coli recA antibodies. This protein increased in abundance 10 fold during purification, similar to the increase in specific activity suggesting that the two are correlated. Using strand exchange method of Rao et al. (1991) and the partially purified recA protein from £. coli we have demonstrated homologous pairing and formation of 3 stranded DNA molecules from ssDNA and homologous dsDNA. However, this recombinase activity has not been found in Chiamvdomona s stromal extracts by performing the same strand exchange assay. This may suggest that Chlamvdomonas chloroplasts have more complex DNA recombination mechanisms than that in £. coli. Although the exact role or roles of this chloroplast recA-like protein in cpDNA recombination and in cpDNA repair pathway is not known, the functional recombinase activity is important to this alga in cpDNA mutation tolerance. 168 5.5 Prospects Transformation of chloroplasts is now possible by the application of microprojectile bombardment for the direct delivery of DNA to chloroplasts in living cells. This allows the study of a much wider range of molecular genetic and biochemical problems involving chloroplasts than before. However, the chloroplast transformation is still in its infancy, especially for higher plants. Further development of higher plant chloroplast transformation probably requires more information about the properties of a chloroplast genome and cell behavior that would be exploitable for the purposes of transformation. In order to gain a better understanding the structure and function of recA-like protein in Ch1amvdomonas chloroplasts, further purification is necessary. The isoelectric focusing gel and HPLC may be considered to be methods for this purpose. The next step would be to determine the sequence of amino acids in the recA-like protein. Then the nucleotide sequence of this protein can be deduced and analyzed. In addition, the analysis of structure and function of recA-like protein will be facilitated as data for mutant recA analogues from Chlamvdomonas or other species become available. Based on the information of sequence of nucleotide and amino acids of recA-like proteins, mutations in the chloroplast recA gene should be made to determine their effects on the cpDNA recombination and repair system in 169 Chlamvdomonas chloroplasts. For example, the recA gene deletion mutant can be used to compare their chloroplast transformation frequency with wild type. It is no doubt that the continued study of the recA analogue will provide us with valuable information for the manipulation of chloroplast genome. REFERENCES

Aguettaz P, Seyer P, Pesey H, Lescure AH (1987) Relations between the plastid gene dosage and the levels of 16s rRNA and rbcL gene transcript, during amyloplast to chloroplast change in mixotrophic spinach cell suspensions. Plant Mol Biol 8:169-177 Angulo JF, Moreau PL, Maunoury R, Laporte J, Hill AM (1989) KIN, a mammalian nuclear protein immunologically related to £. coli recA protein. Mutat Res 217:12 3-134 Backendorf C, Brandsma JA, Kartasova T, van de Putte P (1983) In vivo regulation of the uvrA gene: role of the "-10" and ”-35" promoter regions. Nucleic Acids Res 11:5795- 5810. Baumgartner BJ, Rapp JC, Mullet JE (1989) Plastid transcription activity and DNA copy number increase early in barley chloroplast development. Plant Physiol 89:1011-1018 Bedbrook JR, Smith SM, Ellis RJ (1980) Molecular and of cDNA encoding the precursor to the small subunit of chloroplast ribulose-1,5-bisphosphate carboxylase. Nature 287:692-697 Belliard G, Pelletier G, Vedel F, Quetier F (1978) Morphological characteristics and chloroplast DNA distribution in different cytoplasmic parasexual hybrids of Nicotiana tabacum. Mol Gen Genet 165:231-237 Bennett J, Radcliff C (1975) Plastid DNA replication and plastid division in the garden pea. FEBS Lett 56:222-225 Binding H, Jain SM, Finger FJ, Mordhorst G, Nehis R, Gressel J (1982) Somatic hybridization of an atrazine-resistant biotype of Solanum nigrum with Solanum tuberosum. Theor Appl Genet 63:273-277 Blanar MA, Kneller D, Clark AJ, Karu AE, Cohen FE, Langridge R, Kuntz ID (1984) A model for the core structure of the Escherichia coli recA protein. Cold Spring Harbor Symp Quant Biol 49:507-511

170 171 Blowers AD, Bogorad L, Shark DB, Sanford J (1989) Studies on Ch 1 amvdomonas chloroplast transformation: foreign DNA can be stably maintained in the chromosome. The Plant Cell 1:123-132 Boffey SA, Ellis JR, Sellden G, Leech RM (1979) Chloroplast division and DNA synthesis in light-grown wheat leaves. Plant Physiol 64:502-505 Boffey SA, Leech RM (1982) Chloroplast DNA levels and the control of chloroplast division in light-grown wheat leaves. Plant Physiol 69:1387-1391 Bonnett HT (1976) On the mechanism of uptake of Vaucheria chloroplasts by carrot protoplasts treated with polyethylene glycol. Planta 131:229-233 Bonnett HT, Eriksson T (1974) Transfer of algal chloroplasts into protoplasts of higher plants. Planta 120:71-79 Bonnett HT, Glimelius K (1983) Somatic hybridization in Nicotiana: Behavior of organelles after fusion of protoplasts from male-fertile and male-sterile cultivars. Theor Appl Genet 65:213-217 Boynton JE, Gillham NW, Harris EH, Hosier JP, Johnson AM, Jones AR, Randolph-Anderson BL, Robertson D, Klein TM, Shark KB, Sanford JC (1988) Chloroplast transformation in Chlamvdomonas with high velocity microprojectiles. Science 240:1534-1538 Boynton JE, Gillham NW, Harris EH, Newman SM, Randolph- Anderson BL, Johnson AM, Jones AR (1990) Manipulating the chloroplast genome of Chlamydomonas: and transformation. In M Baltscheffsky ed, Current Research in Photosynthesis, Vol III. Kluwer, Amstredam, pp 509-516 Brenner SL, Mitchell RS, Morrical SW, Neuendorf SK, Schutte BC, Cox MM (1987) RecA protein-promoted ATP hydrolysis occurs throughout recA nucleoprotein filaments. J Biol Chem 262:4011-4016 Brent R, Ptashen M (1981) Mechanism of action of the lexA gene product. Proc Natl Acad Sci USA 78:4204-4208 Bryant FR, Lehman IR (1985) On the mechanism of renaturation of complementary DNA strands by the recA protein of Escherichia coli. Proc Natl Acad Sci USA 82:297-301 172 Bryant FR, Taylor AR, Lehman IR (1985) Interaction of the recA protein of Escherichia coli with single-stranded DNA. J Biol Chem 260:1196-1202 Cannon G, Heinhorst S, Siedlecki J, Weissbach A (1985) Chloroplast DNA synthesis in light and dark grown cultured Nicotiana tabacum cells as determined by molecular hybridization. Plant Cell Reports 4:41-45 Cassuto E, Lightfoot LA, Howard-Flanders P (1987) Partial purification of an activity from human cells that promotes homologous pairing and the formation of heteroduplex DNA in the presence of ATP. Mol Gen Genet 208:10-14 Chow SA, Radding CM (1985) Ionic inhibition of formation of recA nucleoprotein networks blocks homologous pairing. Proc Natl Acad Sci USA 82:5646-5650 Clark AJ (1973) Recombination deficient mutants of E.coli and other bacteria. Ann Rev Genet 7:67-86 Clark AJ, Margulies AD (1965) Isolation and characterization of recombination-deficient mutants of Escherichia coli K12. Proc Natl Acad Sci USA 53:451-459 Costanzo MC, Foe LD (1988) Transformation of yeast by agitation with glass beads. Genetics 12 0:667-67 0 Cotterill SM, Fersht AR (1983) RecA filaments in solution. Biochemistry 22:3525-3531 Cox MM, Lehman IR (1981) Directionality and polarity in recA protein-promoted branch migration. Proc Natl Acad Sci USA 78:6018-6022 Cox MM, Lehman IR (1987) Enzymes of general recombination. Ann Rev Biochem 56:229-2 62 Crossway A, Oakes JV, Irvine JM, Ward B, Knauf VC, Shewmaker CK (1985) Integration of foreign DNA following microinjection of tobacco mesophyll protoplasts. Mol Gen Genet 202:179-185 Cunnigham RP, DasGupta C, Shibata T, Radding CM (1980) Homologous pairing in genetic recombination: recA protein makes joint molecules of gapped circular DNA and closed circular DNA. Cell 20:223-235 Daniell H, Vivekananda J, Nielsen BL, Ye GN, Tewari KK, Sanford JC (1990) Transient foreign gene expression in chloroplasts of cultured tobacco cells after biolistic 173 delivery of chloroplast vectors. Proc Natl Acad Sci USA 87:88-92 De Block M, Schell J, VanMontagu M (1985) Chloroplast transformation by Aqrobacterium tumefaciens. EMBO J 4:1367-1372 de Hass JMf Boot KJM, Haring HA, Kool AJ, Nijkaaup HJJ (1986) A Petunia hybrid chloroplast DNA region, close to one of the inverted repeats, shows sequence homology with the Euglena gracilis chloroplast DNA region that carries the putative replication origin. Mol Gen Genet 202:48-54 de Hass JM, Boot KJM, Haring MA, Kool AJ, Nijlcamp HJJ (1987) Characterization of DNA synthesis and chloroplast DNA replication initiation in a Petunia hybrid chloroplast lysate system. Curr Genet 12:377-386 Dellaporta SL, Wood J, Hinks B (1983) A plant DNA minipreparation: version I. Plant Mol Biol Rep 1:19-21 Deng XW, Gruissem W (1987) Control of plastid gene expression during development: the limited role of transcriptional regulation. Cell 49:379-389 Deno H, Sugiura (1983) The nucleotide sequences of tRNASer (GCU) and tRNAGln (UUG) genes from tobacco chloroplasts. Nucleic Acids Res 11:8407-8414 Deno H, Sugiura M (1984) Chloroplast tRNAGly gene contains a long in the D. stem : Nucleotide sequences of tobacco chloroplast genes for tRNAGly (UCC) and tRNAArg (UCU). Proc Natl Acad Sci USA 81:405-408 Deno H, Kato A, Shinozaki K, Suginra M (1982) Nucleotide sequences of tobacco chloroplast genes for elongator tRNAMe* and tRNAVal (UAC) : the tRNAVal (UAC) gene contain a long intron. Nucleic Acids Res 10:7511-7520 Dunn K, Chrysogelos S, Griffith J (1982) Electron microscopic visualization of recA-DNA filament: evidence for a cyclic extension of duplex DNA. Cell 28:757-765 Eitner G, Adler B, Lanzov VA, Hofemeister J (1982) Interspecies recA protein substitution in Escherichia coli and Proteus mirabilis. Mol Gen Genet 185:481-486 Feinberg AP, Vogelstein B (1983) A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Anal Biochem 132:6-13 174 Flechtner VR, Sager R (1973) Ethidium bromide induced selective and reversible loss of chloroplast DNA. Nature New Biol 241:277-279 Flory J, Radding CM (1982) Visualization of recA protein and its association with DNA: a priming effect of single- strand-binding protein. Cell 28:747-756 Flory J, Tsang SS, Muniyappa K (1984) Isolation and visualization of active presynaptic filaments of recA protein and single-stranded DNA. Proc Natl Acad Sci USA 81:7026-7030 Fromm ME, Taylor LP, Walbot V (1986) Stable transformation of maize after gene transfer by electoporation. Nature 319:791-793 Gellert M, Mizuuch K, O'Dea MH, Iton T, Tomizawa JI (1977) Nalidixic acid resistance: a second genetic character involved in DNA gyrase activity. Proc Natl Acd Sci USA 74:4772-4776 Gellert M, O'Dea MH, Iton T, Tomizawa JI (1976) Novobiocin and coumermycin inhibit DNA supercoiling catalyzed by DNA gyrase. Proc Natl Acad Sci USA 73:4474-4478 Geoghegan CM, Houghton JA (1987) and isolation of a cyanobacterial gene which increases the uv and methyl methanesulphonate survival of recA strains of Escherichia coli K12. J Gen Microbiol 13:119-126 Giles KL, Vaughn V, Ranch JP, Emery J (1980) Liposome-mediated uptake of chloroplasts by plant protoplasts. In Vitro 16:581-594 Gleba Y, Komarnitshy IK, Kolesnik NN, Meshkene I, Martyn GI (1985) Transmission genetics of the somatic hybridization process in Nicotiana. II Plastone heterozygotes. Mol Gen Genet 198:476-481 Gold B, Carrillo N, Tewari KK, Bogorad L (1987) Nucleotide sequence of a preferred maize chloroplast genome template for in vitro DNA synthesis. Proc Natl Acad Sci USA 84:194-198 Gonda DK, Radding CM (1983) By searching processively recA protein pairs DNA molecules that share a limited stretch of homology. Cell 34:647-654 Gonda DK, Radding CM (1986) The mechanism of the search for homology promoted by recA protein. J Biol Chem 261:13087- 13096 175 Harris EH (1989) The Chlamvdomonas source book. Academic Press, New York Harris EH, Boynton J, Gillham N (1987) Genetic Haps. Vol 4. Cold Spring Harbor pp257-277 Heinhorst S, Cannon G, Weissback A (1985) Chloroplast DNA synthesis during the in cultured cells of Nicotiana tabacum: Inhibition by nalidixic acid and hydroxyurea. Arch Biochem Biophys 239:475-479 Heinhorst S, Cannon G, Weissback A (1986) DNA synthesis in suspension cultures of Nicotiana tabacum and Glvcine max. In Akoyunoglou G, Senger H (eds) Regulation of chloroplast differentiation. Alan R Liss Inc, New York pp 447-487 Hirschberg J, McIntosh L (1983) Molecular basis of herbicide resistance in Amaranthus hvbridus. Science 222:1346-1349 Herman PG, Possingham JV (1980) In: Reinert J (ed) Results and Problems in Cell Differentiation. Vol 10, pp 45-85 Horii T, Ogawa T, Ogawa H (1980) Organization of the recA gene of Escherichia coli. Proc Natl Acad Sci USA 77:313-317 Hosier JP, Wurtz EA, Harris EH, Gillham NW, Boynton JE (1989) Relationship between gene dosage and gene expression in the chloroplast of Chlamvdomonas reinhardtii. Plant Physiol 91:648-655 Howard-Handers P, Weste SC, Stasiak A (1984) Role of recA protein spiral filaments in genetic recombination. Nature 309:215-220 Hsieh P, Meyn JMS, Camerini-Otero RD (1986) Partial purification and characterization of a recombinase from human cells. Cell 44:885-894 Inamine G, Nash B, Brotn WH (1985) Light regulation of the synthesis of the large subunit of ribulose-1,5- bisphosphate carboxylase in peas: Evidence for translational control. Proc Natl Acad Sci USA 82:5690- 5694 Kato A, Skhimada H, Kusuda M, Sugiura M (1981) The nucleotide sequences of two tRNAAsn genes from tobacco chloroplasts. Nucleic Acids Res 9:5601-5607 Kato A, Takaiwa F, Shinozaki K, Sugiura M (1985) Location and nucleotide sequence of the genes for tobacco chloroplast tRNAAr® (ACG) and tRNALou (UAG) . Curr Genet 9:405-409 176 Keener SL, McNamee KP, McEntee K (1984) Cloning and characterization of recA genes from Proteus vulgaris. Erwinia carotovora, shigella flexneri, and Escherichia coli B/r. J Bacteriol 160:153-160 Kenne K, Lindquist S (1984) A DNA-recombinogenic activity in human cells. Nucleic Acids Res 12:3057-3068 Kenyon CJ, Walker GC (1980) DNA-damage agents stimulate gene expression at specific loci in Escherichia coli. Proc Natl Acad Sci USA 77:2819-2823 Kenyon CJ, Brent R, Ptashne M, Walker GC (1982) Regulation of damage-inducible genes in Escherichia coli. J Mol Biol 160:445-457 Kindle KL (1990) High-frequency nuclear transformation of Chlamvdomonas reinhardtii. Proc Natl Acad Sci USA 87:1228-1232 Kindle KL, Richards KL, Stern DB (1991) Engineering the chloroplast genome: Techniques and capabilities for chloroplast transformation in Chlamvdomonas reinhardtii. Proc Natl Acad Sci USA 88:1721-1725 Klein TM, ED, Wu R, Sanford JC (1987) High-Velocity microprojectiles for delivering nucleic acids into living cells. Nature 327:70-73 Kmiec EB, Holloman WK (1982) Homologous pairing of DNA molecules promoted by a protein from Ustilaao. Cell 29:367-364 Kmiec EB, Hollomwn WK (1984) Synapsis promoted by Ustilaao reel protein, cell 36:593-598 Knight KL, McEntee K (1985a) Covalent modification of the recA protein from Escherichia coli with the phytoaffinity label 8-azidoadenosine 5 1-triphosphate. J Biol Chem 260:867-872 Knight KL, McEntee K (1985b) Affinity labeling of a residue in the ATP binding site of the recA protein from Escherichia coli with 5'-p-fluorosulfonyl- benzoyadenosine. J Biol Chem 260:10177-10184 Roller B, Delins H (1982) Origin of replication in chloroplast DNA of Euglena gracilis located close to the region of variable size.EMBO J 1:995-998 177 Kolodner RD, Tewarl KK (1975) Chloroplast DNA from higher plants replicates by both the Cairns and the rolling circle mechanism. Nature 256:708-711 Konforti BB, Davis RW (1987) 3' Homologous free ends are required for stable joint molecule formation by recA and single-stranded binding proteins in Escherichia coli. Proc Natl Acad Sci USA 84:690-694 Kowalczykowski SC, Clow J, Krupp RA (1987) Properties of the duplex DNA-dependent ATPase activity of Escherichia coli recA protein and its role in branch migration. Proc Natl Acad Sci USA 84:3127-3131 Kucherlapati RS, Spencer J, Moore PD (1985) Homologous recombination catalyzed by human cell extracts. Mol Cell Biol 5:710-720 Lam E, Chua NH (1987) Chloroplast DNA gyrase and in vitro regulation of transcription of template topology by novobiocin. Plant Mol Biol 8:415-424 Lamppa G, Bendick A (1979) Changes in chloroplast DNA levels during development of pea fPisum sativaui. Plant Physiol 64:126-130 Laulhere JP, Rozier C (1976) One-step extraction of plant nucleic acids. Plant Mol Biol 8:415-424 Lawrence ME, Possingham JW (1986) Microspectrofluorometric measurement of chloroplast DNA in dividing and expanding leaf cells of Spinacea oleracea. Plant Physiol 81:708-710 Lewin B (1990) Genes IV. Oxford University Press. Oxford. Little JW (1984) Autodigestion of lexA and phage lambda repressers. Proc Natl Acad Sci USA 81:1375-1379 Little JW, Mount DW (1982) The SOS regulatory system of Escherichia coli. Cell 29:11-22 Little JW, Edmiston SH, Pacelli LZ, Mount DW (1980) Cleavage of the Escherichia coli lexA protein by the recA protease. Proc Natl Acad Sci USA 77:3225-3229 Lockshon D, Morris DR (1983) Positively supercoiled plasmid DNA is produced by treatment of Escherichia coli with DNA gyrase inhibitors. Nucleic Acids Res 11:2999-3017 Lorz H, Baker B, Schell J (1985) Gene transfer to cereal cells mediated by protoplast transformation. Mol Gen Genet 199:178-182 178 Lovett CMJr, Roberts JW (1985) Purification of a RecA protein analogue from . J Biol Chem 260:3305- 3303 Lyman H (1967) Specific inhibition of chloroplast replication in Euolena gracilis by nalidixic acid. J Cell Biol 35:726-730 Lyman H, Jupp AS, Larrinua I (1975) Action of nalidixic acid on chloroplast relocation in Eualena gracilis. Plant Physiol 55:390-392 Maniatis T, Fritsch EF, Sambrollk J (1982) Molecular cloning: A laboratory manual. Cold Spring Harbor, Cold Spring Harbor Laboratory Press McEntee K (1985) Kinetics of DNA renaturation catalyzed by the recA protein of Escherichia coli. Biochemistry 24:4245- 4351 McEntee K, Weinstock GM, Lehman IR (1979) Initiation of general recombination catalyzed in vitro by the recA protein of Escherichia coli. Proc Natl Acad Sci USA 76:2615-2619 McEntee K, Weinstock GM, Lehman JR (1981) Binding of the recA protein of Escherichia coli to single- and double­ stranded DNA. J Biol Chem 256:8835-8844 Medgyesy P, Fejes E, Maliga P (1985) Interspecific chloroplast recombination in a Nicotiana somatic hybrid. Proc Natl Acad Sci USA 82:6960-6964 Medgyesy P, Gollong R, Nagy F (1985) A light sensitive deficient for the effective transfer of chloroplast and mitochondria traits by protoplast fusion in Nicotiana. Theor Appl Genet 70:590-594 Medygesy P, Mewczel L, Maliga P (1980) The use of cytoplasmic streptomycin resistance: chloroplast transfer from Nicotiana tabacum into Nicotiana svivestvis. and isolation of their somatic hybrids. Mol Gen Genet 179:693-698 Meinkoth J, Wahl G (1984) Hybridization of nucleic acids immobilized on solid supports. Anal Biochem 138:267-284 Menczel L, Nagy F, Kiss ZR, Maliga P (1981) Streptomycin resistant and sensitive somatic hybrids of Nicotiana tabacum + Nicotiana knightiana+: correlation of resistance to N. tabacum plastids. Theor Appl Genet 59:191-195 179 Menetski JP, Kowalxzykowski SC (1985) Interaction of recA protein with single-stranded DNA quantitative aspects of binding affinity modulation by nucleotide cofactors. J Mol Biol 181:281-295 Menetski JP, Kowalczykowski SC (1987) Transfer of recA protein from one polynucleotide to another. J Biol Chem 262:2085- 2092 Miyamura S, Nagata T, Kuroiwa T (1986) Quantitative fluorescence microscopy on dynamic changes of plastid during wheat development. Protoplasma 133:66-72 Morrical SW, Cox MM (1985) Light Scattering studies of the recA protein of Escherichia coli: Relationship between free recA filaments and the recA-ssDNA complex. Biochemistry 24:76-767 Mullet JE (1988) Chloroplast development and gene expression. Ann Rev Plant Physiol Plant Mol Biol 39:475-502 Mulligan B, Schultes N, Chen L, 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 Murashige T, Skoog F (1962) A revised medium for rapid growth and bioassay with tobacco tissue cultures. Physiol Plant 15:473-497 Murphy RC, Bryant DA, Porter RD, de Narsac NT (1987) Molecular cloning and characterization of the recA gene from the cyanobacterium Svnechococcus sp. strain PCC 7002. J Bacteriol 169:2739-2747 Murphy RC, Gasparich GE, Bryant DA, Porter RD (1990) Nucleotide sequence and further characterization of the Svnechoccus so. strain PCC7002 recA gene: complimentation of a cyanobaacterial recA mutation by the Escherichia coli recA gene. J Bacteriol 172: Murray MG, Thompson WF (1980) Rapid isolation of high molecular weight plant DNA. Nucleic Acids Res 8:4321-4325 Nitsch JP, Nitsch C (1969) Haploid plants from grain. Science 163:85-87 Ogawa T, Wabiko H, Tsurimoto T, Horii T, Masukata H, Ogawa H (1978) Characteristics of purified recA protein and the regulation of its synthesis in vivo. Cold Spring Harbar Symp Quant Biol 43:909-915 180 Ohme M, Kamogashira T, Shinozaki K, Sugiura M (1984) Locations and sequences of tobacco chloroplast genes for tRNAPro (UGG) , tRNATrp, tRNAfHet and tRNAcly (GCC) : the tRNAGly contains only two base-pairs in D stem. Nucleic Acids Res 12:6741-6749 Ohme M, Kamogashira T, Shinozaki K, Sugiura M (1985) Structure and cotranscription of tobacco chloroplast genes for tRNAGlu (UUC) # tRNATyr (GUA) and tRNAA,ip (GUC) . Nucleic Acids Res 13:1045-1046 Ohyama K, Fukuzawa H, Kohchi T, Shirai H, Sano T, Sano S, Umesoneo K, Shiki Y, Takeuchi M, Chang Z, Auta S, Inokuchi H, Oaeki H (1986) Chloroplast gene organization deduced from complete sequence of liverwort Marchantia polymorph chloroplast DNA. Nature 322:572-574 Ohtani T, Uchimiya H, Kato A, Harada H, Sugita M, Sugiura M (1984) Location and nucleotide sequence of a tobacco chloroplast DNA segment capable of replication in yeast. Mol Gen Genet 195:1-4 Oppermann T, Hong TH, Surzycki SJ (1989) Chloroplast and nuclear genomes of Chlamvdomonas reinhardtii share homology with Escherichia coli genes for DNA replication, repair and transcription. Curr Genet 15:39-4 8 Owttrim GW, Coleman JR (1987) Molecular cloning of a recA-like gene from the cyanobacterium Anabaena variabilis. J Bacteriol 169:1824-1829 Owttrim GW, Coleman JR (1989) Regulation of expression and nucleotide sequence of the Anabaena variabilis recA gene. J Bacteriol 171:5713-5719 Pelletier G, Primard C, Vedel F, Chetrti P, Pemy R, Rousselle, Remard M (1983) Intergeneric cytoplasmic hybridization in Cruciferae by protoplast fusion. Mol Gen Genet 191:244- 250 Peterson KR, Mount DW (1987) Differential repression of SOS genes by unstable lexA 41 (Tsl-1) protein causes a "split-phenotype" in Escherichia coli K-12. J Mol Eiol 193:27-40 Pienkos P, Waldield A, Hershberger CL (1974) Effect of nalidixic acid on Euolena gracilis: Induced lose of chloroplast deoxyribonucleic acid. Arch Biochem Biophys 165:548-553 Possingham JV, Lawence ME (1983) Controls to plastid division. Int Rev Cytol 84:1-56 181 Radman M (1975) SOS repair hypothesis: phenomenology of an inducible DNA repair which is accompanied by mutagenesis.In Molecular Mechanisms for Repair of DNA. Hanawalt P, Setlow RB (eds) New York: Plenum Press pp:355-367 Rao BJ, Duterix M, Radding CM (1991) Stable three-stranded DNA made by recA protein. Proc Natl Acad Sci USA 88:2984-2988 Ravel-chapuis Pf Heizman P, Nivon V (1982) Electron microscopic localization of the replication origin of Eualena gracilis chloroplast DNA. Nature 300:78-81 Register JC, Griffith J (1985) The direction of recA protein assembly onto single strand DNA is the same as the direction of strand assimilation during strand exchange. J Biol Chem 260:12308-12312 Roberts JW, Roberts CW, Graig NL, Phizicky EM (1978) Activity of the Escherichia coli recA-gene product. Cold Spring Harbor Symp Quant Biol 43:917-920 Rochaix JD, van Dillewijn J, Rahire M (1984) Construction and characterization of autonomously replicating plasmids in the green unicellular alga Chlamvdomonas reinhardtii. Cell 36:925-931 Roffey R, Golbeck JH, Hille CR, Sayre R (1991) Photosynthetic electron transport in genetically altered photosystem II reaction centers of chloroplast. Proc Nat Acad Sci USA. In press. Rusche JR, Konigsberg W, Howard-Flanders P (1985) Isolation of altered recA polypeptides and interaction with ATP and DNA. J Biol Chem 260949-955 Salles B, Paoletti C (1983) Control of UV induction of recA protein. Proc Natl Acad Sci USA 80:65-69 Salles B, Lang MC, Freund AM, Paoletti L, Daune M, Fuchs RP (1983) Different levels of induction of recA protein in £• coli (PQ10) after treatment with two relate . Nucleic Acids Res 11:5235-5242 Sancar A, Sancar DB (1988) DNA repair enzymes. Annu Rev Biochem 57:29-67 Sancar GB, Sancar A, Little JW, Rupp WD (1982) The uvrB gene of Escherichia coli has both lexA-repressed and lexA- independent promoters. Cell 28:523-530 182 Sancar A, Stachelek C, Konigsberg W, Rupp HD (1980) Sequences of the recA gene and protein. Proc Natl Acad Sci USA 77:2611-2615 Sanford JC, Klein TM, Wolf ED, Allen N (1987) Delivery of substances into cells and tissues using a particle bombardment process. Particulate Sci Technol 5:27-37 Sasaki Y, Nakamura Y, Matsuno R (1986) Phytochrome-mediated accumulation of chloroplast DNA in pea leaves. FEBS Letters 196:171-174 Sayre RT, Andersson B, Bogorad L (1986) The topology of a : the orientation of the 32 kD Qb-binding chloroplast thylakoid membrane protein. Cell 47:601-608. Schlunegger B, Fasnacht M, Stutz E, Koller B, Delius H (1983) Analysis of a polymorphic region of the Eualena gracilis chloroplast genome. Biochem Biophys Acta 739:114-121 Schlunegger B, Stutz E (1984) The Eualena araeilis chloroplast genome: structural features of a DNA region possibly carrying the single origin of DNA replication. Curr Genet 8:629-634 Shibata T, Cunningham RP, DasvGupta C, Radding CM (1979) Homologous pairing in genetic recombination: complexes of recA protein and DNA. Pore Natl Acad Sci USA 76:5100-5104 Shibata T, Makino O, Ikawa S, Ohtani T, Iwabuchi N, Shibata Y, Maaeda H, Audo T (1984) Roles of processive unwinding in recombination reactions promoted by recA protein of Escherichia coli: a study using a monoclonal antibody. Cold Spring Harbor Symp Huant Biol 49:541-551 Shimada H, Sugiura M (1991) Fine structural features of the chloroplast genome: comparison of the sequenced chloroplast genomes. Nucleic Acids Res 19:98 3-995 Shinozaki K, Ohme M, Tanaka T, Wakasugi T, Hayashida N, Matsubayahi T, Zaita N, Chunwongse J, Oboracawa K, Meng BY, Sugita M, Deno H, Kamogashira T, Yamada K, Kusnda J, Takaium, Kato A, Tohcloh N, Shimada H, Sugiura M (1986) The complete nucleotide sequence of the tobacco chloroplast genome: its gene organization and expression. EMBO J 5:2043-2049 Slilaty SN, Rupley JA, Little JH (1986) Intramolecular cleavage of lexA and phage lambda regressors: dependence of kinetics on repressor concentration, pH, temperature and solvent. Biochemistry 25:6866-6875 183 Smith 6R (1987) Mechanism and control of homologous recombination in Escherichia coli. Ann Rev Genet 21:179- 201 Souther EM (1975) Detection of specific sequences among DNA fragments separated by gel electrophoresis. J Mol Biol 98:503-517 Stirdivant SM, Crossland LD, Bogorad L (1985) DNA supercoiling affects in vitro transcription of two maize chloroplast genes differently. Proc Natl Acad Sci USA 82:4886-4890 Strittmattter G, Gozdzicka-Jozediak A, Kossel H (1985) Identification of an rRNA promoter from Zea mavs chloroplasts which excludes the proximal tRNAv“GAC from the primer transcript. EMBO J 4:599-604 Sueoka N (1960) Mitotic replication of deoxyribonucleic acid in Chlamvdomonas reinhardtii. Proc Natl Acad Sci USA 46:83-91 Sugino A, Peebles CL, Kreuzer KN, Cozzarelli NR (1977) Mechanism of action of nalidixic acid: purification of Escherichia coli nalA gene product and its relationship to DNA gyrase and a novel nicking-closing enzyme. Proc Natl Acad Sci USA 74:4767-4774 Sugino A, Higgins NP, Brown PO, Peebles CL, Cozzarelli NR (1978) Energy coupling in DNA gyrase and the mechanism of action of novobiocin. Proc Natl Acad Sci USA 75:4838- 4842 Sugita M, Kato A, Shimada H, Sugiura M (1984) Sequence analysis of the junctions between a large inverted repeat and single-copy regions in tobacco chloroplast DNA. Mol Gen Genet 194:200-205 Sugita M, Shinozaki K, Sugiura M (1985) Tobacco chloroplast tRNALy8(UUU) gene contains a 2.5-kilobase pair intron: An open reading frame and a conserved boundary sequence in the intron. Proc Natl Acad Sci USA 82:3557-3561 Surzycki S (1969) Genetic functions of the chloroplast of Chlamvdomonas reinhardtii: Effect of rifampicin on chloroplast DNA dependent RNA polymerase. Proc Natl Acad Sci USA 63:1327-1334 Svab Z, Hajdukiewicz P, Maliga P (1990) Stable transformation of plastids in higher plants. Proc Natl Acad Sci USA 87:8526-8530 184 Towhin H, Staehlin E, Gordon K (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets. Procedures and some applications Proc Natl Acad Sci 76:4250-4354 Thomas B, Wachwenagel W (1987) Regulatory role of recF in the "SOS" response of £. coli; impaired induction of nS0S" genes by UV irradiation and nalidixic acid in a recF mutant. J Bacteriol 169:1731-1736 Thompson WF, Everett M, Polans NO, Jorgensen RA, Palmer JD (1983) Phytochrome control of RNA levels in developing pea and mung-bean leaves. Planta 158:487-500 Tsang SS, Chow SA, Radding CM (1985) Networks of DNA and recA protein are intermediates in homologous pairing. Biochemistry 24:3226-3232 Tsang SS, Muniyappa K, Axhderian E, Gonda DK, Radding CM, Flory J, Chase JH (1985) Intermediates in homologous pairing promoted by recA protein. J Mol Biol 185:295-309 Uchimiya H,m Ohtani T, Ohgawara T, Harada H, Sugita M, Sugiura M (1983) Molecular cloning of tobacco chromosomal and chloroplast DNA segments capable of replication in yeast. Mol Gen Genet 192:1-4 Vallet JM, Rochaiz JD (1985) Chloroplast origins of DNA replication are distinct from chloroplast ARS sequences in two . Current Genetics 9:321-324 WaddellJ, Wang ZM, Wu M (1984) Electron microscopic localization of the chloroplast DNA replicative origins in Chlamvdomonas reinhardtii. Nucleic Acids Res 12:3843- 3856 Wakasugi T, Ohme M, Shinozaki K, Sugiura M (1986) The 5'part of the gene for S12 is located 30 kbp downstream from its 3* part in tobacco chloroplast genome. Nucleic Acids Res 14:3243 Wang XM, Chang CH, Waddell J, Wu M (1984) Cloning and delimiting one chloroplast DNA replicative origin of Chlamydomonas. Nucleic Acids Res 12:3857-3872 Watt WM, Ingles CJ, Urdea MS, Rutter WJ (1985) Homology requirements for recombination in Escherichia coli. Proc Natl Acad Sci USA 82:4768-4772 Wehrli E, Knusel F, Schmid K, Staekhlin M (1968) Interaction of rifampicin with bacterial RNA polymerase. Proc Natl Acad Sci USA 61:667-673 185 Weinstock GM, McEntee K, Lehman IR (1979) ATP-dependent renaturation of DNA catalyzed by the recA protein of Escherichia coli. Proc Natl Acad Sci USA 76:126-130 Weinstock GM, McEntee K, Lehman IR (1981) Hydrolysis of nucleoside triphosphates catalyzed by the recA protein of Escherichia coli. J Biol Chem 256:8829-8834 Well JH (1987) Organization and expression of the chloroplast genome. Plant Science 49:149-157 Whitfield PR, Bottomley W (1983) Organization and structure of chloroplast genes. Ann Rev Plant Physiol 34:279-310

Willmitzer L (1988) The use of transgenic plants to study plant gene expression. TIG 4:13-18 Witkin EM (1976) Ultraviolet mutagenesis and inducible DNA repair in Escherichia coli. Bacteriol Rev 40:867-907 Wu N, Lou JK, Chang DY, Chang CH, Nie ZQ (1986) Structure and function of a chloroplast DNA replication origin of Chlaroydomonas reinhardtii. Proc Natl Acad Sci USA 83:6761-6765 Wurtz EA, Boynton JE, Gillham NW (1977) Perturbation of chloroplast DNA amounts and chloroplast gene transmission in Chlamvdomonas reinhardtii by 5-fluorodeoxyuridine. Proc Natl Acad Sci USA 74:4552-4556 Wurtz EA, Sears BB, Robert DK, Shpherd HS, Gillham NW, Boynton JE (1979) A specific increase in chloroplast gene mutation following growth of Chlamvdomonas in 5- fluorodeoxyuridine. Mol Gen Genet 170:235-242 Yamada K, Shimoraki K, Sugiura M (1986) DNA sequences of tobacco chloroplast genes for tRNASer (GGA) , tRNA Thr (UGU) , tRNALeu (UAA) , tRNAphe (GAA) : the tRNALeu gene contains a 503 bp intron. Plant Mol Biol 6:193-199 Yasuda T, Kuroiwa T, Nagata T (1988) Preferential synthesis of plastid DNA and increased replication of plastids in cultured tobacco cells following medium renewal. Planta 174:235-241 Zimmermann U, Vienken J (1982) Electric field-induced cell-to- cell fusion. J Memb Bio 67:165-182.