ll’l'l..lt'.- . LIBRARY Michigan State University This is to certify that the thesis entitled Studies of transmission and recombination of Oenothera plastids presented by Wan-Ling Chiu has been accepted towards fulfillment of the requirements for Master degree in Science Major professor Damage data ”5’27 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution MSU LIBRARIES ._cI—. RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. STUDIES OF TRANSMISSION AND RECCMBINATION OF 0W PLASTOMES BY wan-Ling Chiu A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Botany and Plant Pathology 1987 ABSTRACT STUDIES OF TRANSMISSION AND RECOMBINATION OF OEWUTHEHM PLASTOMES By Wan-Ling Chiu The likelihood of genetic recombination between higher plant chloro— plasts was examined using Obnothera, a genus which exhibits biparental plastid transmission. Recombination was not detected among 10,000 progeny from crosses between plants carrying different plastid mutants. The transmission abilities of the four Ganotbera plastome types were compared in a constant nuclear background. In crosses between plants carrying mutant and wild-type plastids, the frequency of variegated progeny ranged from zero to 56%. From this frequency and the extent of variegation, it is apparent that both the efficiency and the timing of onset of plastid multiplication after fertilization are plastome- dependent. Three autonomous replication sequences from the Chlamydbnanas chloroplast genome were used to probe for homologous sequences in Oenothera chNA. Differences in hybridization patterns exist among dif- ferent plastome types. An attempt was also made to test if cyanobacteria can serve as a heterologous system for the functional analysis of chNA origins of replication. This thesis is dedicated to Drs. W. Stubbe and F. Schotz for their delicate work on Oenotbera. ii ACKNOWLEDGMENT I would like to thank my major advisor Dr. Barbara Sears for introducing me to the world of organelle genetics and Genothera. I appreciate Dr. J. Hancock and Dr. L. McIntosh for being on my thesis committee. Special thanks should also go to Dr. N. E. Good for his long term encouragement, Dr. Linda Schnabelrauch for her advice in plant tissue culture, and all my friends in the lab for their help in the field work. Finally, I thank JAE and WSCE for their support in every respect. iii TABLE OF CONTENTS page LIST OF TABLES ....................................................... vi LIST OF FIGURES ..................................................... vii manION ...... .00... ....... 00.0000...OOOOOOOOOOOOOOO. ....... 00.0.01 CHAPTER 1. RECOMBINATION BETWEEN CRLOROPLAST DNAs DOES NOT OCCUR IN CROSSES 05' 0mm ....................................... ‘7 Introduction ....................................................... 7 Materials and.Methods .............................................. 9 Plant Material..... ..... . ...................................... 9 Crosses and Selection of Presumptive Recombinants ............... 9 Isolation of Chloroplast DNA ................................... 11 Analysis of Chloroplast DNAs ................................... 11 Results ........................................................... 11 Selection of Presumptive Plastome Recombinants ................. 11 Chloroplast DNA Analysis ....................................... 14 Discussion ........... . .................................... . ....... 20 CHAPTER 2. PLASTID INRERITANCE IN dfiflflflfififiu: ORGANELLE GENOME INFLUENCES THE EXTENT OF BIPARENTAL TRANSMISSION .......... 24 Introduction ..... .... ............................................. 24 Materials and Methods ............................................. 27 Plant Material and Crosses ..................................... 27 Germination and Scoring of the Seedlings ....................... 29 Results ........................................................... 29 White x Green Crosses ...................................... ....30 Green x White Crosses .......................................... 37 Statistical Analysis of Biparental Transmission Frequencies....37 Ccmparisons of Different Mutant Plastids ....................... 42 iv Discussion ...................... ...... ........... . ................ 42 CHAPTER 3. INDIRECT APPROACHES TO CHARACTERIZE CHLOROPLAST DNA ORIGENS OF REPLICATION .................................... 52 Introduction ......... . ............................................ 52 Materials and Methods .............. ............ ................... 56 Isolation of Chloroplast DNA from Chlamydhmonas. ......... ......56 DNA Blot Hybridization ......................................... 57 Cloning of the Chloroplast DNA Origin of Replication from Chlamydhlonas. ............................... ......... ..68 Bacterial Strains and Growth Conditions........................58 Conjugation of Plasmids into Cyanobacteria.....................61 Transformation of synecbocystjs'PCC 6803 ....................... 62 Isolation of Total DNA from symecbocystjs'PCC 6803 .......... ...62 ResultSOOOOOOOOOOOO0.00.0...00......OOOOOOOOIOOOOOOOOOO0.0.00.00.063 Detection of sequences homologous to ARCs in canotbera chNA...63 Testing autonomous replication activity of ariA in cyanobacteria............................... ........... ..68 Conjugation of an arid-containing Plasmid into Cyanobacteria. ............................. . ............. 68 Transformation of Cyanobacteria with oriA-containing plasmids ........ . ....... . ................................ 70 Discussion ..... . ......................... . ........................ 73 CONCLUSION ........................................................... 77 LIST OF REFERENCES ................................................... 81 LIST OF TABLES page Table 1. Phenotypes of progeny from crosses of plastome mutants (first season) .............................................. 13 Table 2. Phenotypes of progeny from crosses of plastome mutants (second season) ............... ........ ................ . ..... 19 Table 3. Biparental transmission of chloroplasts in White x Green crosses ..................................................... 31 Table 4. Biparental transmission of chloroplasts in Greeg43 White crosses ........ ... ............ ................ ..... .. ....... 38 Table 5. Analysis of variance of biparental transmission frequencies... .................... ...... .................... 41 Table 6. Transformation of synecbocysth'PCC 6803 .................... 72 vi LIST OF FIGURES page Figure 1: CpDNA restriction fragment patterns ........................ 16 Figure 2: Frequency of biparental transmission in White x Green crusaoossooososoassessosssooosoossososs00000000000000.00034 Figure 3: Abundance of tissue containing paternal plastids in‘ variegated progeny from White x Green crosses.... .......... 36 Figure 4: Frequency of biparental transmission in Green x White crosses ...... . ........... . ................................. 40 Figure 5: Maps of the plasmids carrying 6L reinbardfij chNA arid ..... 60 Figure 6: Hybridization of ARCl element to Genotbera'chNA...........65 Figure 7: Hybridizations of ARCZ and ARCSb elements to Cenotbera chNA.. .............................. . ..................... 67 vii INTRODUCTION The plastid is an essential organelle for the normal function of plant cells. It is responsible for aspects of amino acid biosynthesis and fatty acid synthesis in addition to photosynthesis (reviewed by Kirk and Tilney-Bassett, 1978). The genome sizes of plastids range from 120 to 217 kb (reviewed by Palmer, 1985). Genes for about 50 chloroplast proteins, 4 different rRNAs, and 30 different tRNAs have been located on the chloroplast genome (Shinozaki et 81., 1986). Sequencing of the entire plastid genome (plastome) of tdbacco (Shinozaki at 81., 1986) and liverwort (Ohyama et a1., 1986) has revealed a number of large unidentified open reading frames. Thus, the functions of many genes on the plastome remain to be determined. Nevertheless, the great majority of genes for chloroplast proteins are nuclear-encoded (Kirk and Tilnebeassett, 1978). Hence, the development and the normal function of chloroplasts require a close interaction be- tween the genes of the two compartments. Most studies concerning the in- teraction between the nuclear and plastid compartments have concentrated on the development of photosynthetically active chloroplasts. It is known that the mutation of genes encoding chloroplast components in one of the two genomes can block the expression of genes in the other (e.g. Mascia and Robertson, 1978). A physiological disturbance of the chloroplast can also affect the expression of nuclear genes encoding chloroplast proteins (Burgess and Taylor, 1987; Mayfield and Taylor, 1987). The close regulation of the interaction between the two genomes is especially obvious when chloroplast genome is placed in a fereign nuclear background. A phenomenon called hybrid bleaching caused by in- compatabilities between nuclear and chloroplast genomes is well docue mented in the genera of 0enotbera (Stubbe, 1959) and Pelargonim (Metzlaff et 81., 1982). Since plastids are indispensible parts of a plant cell, some mechanism is required to ensure the presence of at least one plastid per cell during cell growth and division (reviewed by Boffey, 1984; Possingham, '1984). Clearly, the precision of this process is not equivalent to that of the nucleus, where DNA replicates only once during the cell cycle and where the mitotic spindle assures an equal partition of the genetic material. Experiments using inhibitors of protein synthesis indicate that nuclear gene products are required for both chloroplast division and chNA syn- thesis (Hashimoto and Murakami, 1983; Reinhorst et 81., 1985). In- hibitors of chloroplast protein synthesis affect these two cellular processes to a very limited extent. The presence of chNA in chloroplast mutants lacking ribosomes also supports the concept that the processes of plastid multiplication are mainly controlled by the nuclear genome (Scott et a1., 1982). On the other hand, Schfitz (1954, 1974) studied plastid transmission in the genus Cenotbera'and suggested that the plas- tome itself determines the intrinsic multiplication rates of the plastid (reviewed by Kirk and Tilney-Bassett, 1978). According to these studies, both nuclear and plastid genomes can affect the process of plastid multiplication. The interactions between these two compartments in plastid multiplication have not been directly studied, although they must exist. The characterization of mutations is a basic genetic technique for the identification of gene function. A large number of chloroplast mutants have been isolated (reviewed by Borner and Sears, 1986). Unfortunately, the characterization of plastid mutants has yielded only limited infor- mation regarding the function of plastid genes. The site of the genetic lesion is generally difficult to locate, and many of the chloroplast mutations have the same chlorophyll-deficient phenotype and lack photosynthetic activity. The methods used for characterizing mutations in prokaryotes have not been applied to chloroplasts for several reasons. First of all, there is no reliable means yet available to genetically transform the chloroplast. Thus, wild-type or mutant DNA of a chloroplast cannot be returned to the chloroplast to directly test its function. Secondly, recombination between different plastomes has not been observed to occur naturally except in the sexual crosses of the unicellular green alga, Chlamydhmonas’ (Gillham, 1978). Therefore, a mutation of a higher plant plastome can neither be characterized by com- plementation with a’ wild-type gene nor through classical genetic mapping. For these reasons, many gene functions in chloroplasts have been studied only in model systems such as unicellular algae or photosynthetic bacteria. The genus Genotbera'is uniquely suited as a system for the study of nuclear—chloroplast interactions with regards to. both chloroplast development and plastid multiplication. First, plastids of Oenotbera are transmitted from both parents (biparental transmission) in. sexual crosses (reviewed by Kirk and Tilney-Bassett, 1978). Hence, two types of plastids can be brought into the same cell to allow the study of their interactions and to provide an opportunity for recombination events be- tween their plastomes. Second, broad interspecific crosses are possible (Reviewed by Cleland, 1972). The interactions between plastid and nuclear genomes can be studied in a wide variety of plastome-genome combinations. Third, due to extensive chromosomal translocations, cer- tain haploid genomes of Genotbera are transmitted as intact complexes during crosses (reviewed by Cleland, 1972). This facilitates the rapid movement of plastids into a desired nuclear background (Stubbe and Harrmann, 1982). Plastome-genome incompatability and differential plas- tid transmission in this genus have both been well documented (reviewed by Kirk and Tilnebeassett, 1978). However, in neither case have the the phenomena been characterized at the molecular level. The experiments in the first chapter of this thesis were designed to recover wild-type plastome recombinants from crosses between Cenotbera plants carrying only mutant plastids in their germ layers. The underly— ing motive for this line of experimentation was the desire to associate regions of the plastome DNA with certain plastid-determined traits, in particular, plastome-genome incompatability and differential plastid transmission. In order to understand the cooperative interaction between the two genetic compartments in the process of plastid multiplication, the roles of plastid and nuclear gems-es should be examined more carefully. In Chapter 2 of this thesis, the transmission abilities of different plas- tome types of Oenotbera were examined in a constant nuclear background to exclude the effect of nuclear variability and to concentrate ini- tially on the role of plastome itself in this process. Much of the data concerning Genotberajplastid transmission can be ex- plained by postulating that the plastome exerts control over the multi- plication of the plastid (Schétz, 1974). However, no plastid gene products appear to be essential to this process (Scott et a1., 1982). This apparent inconsistency could be resolved if some chloroplast DNA structure, such as the origin of chNA replication, were the major plas- tid determinant for the differential plastid transmission observed in Genothera. Chapter 3 of this thesis presents some preliminary efforts towards assessing this hypothesis. DNA hybridization was used in an at- tempt to see if the frequency and the distribution of chNA sequences that may be important for chNA replication correlates with the various transmission abilities of Oenotbera plastids. In order to be able to analyze the functional requirements for a chNA origin of replication, several heterologous systems were tested for their abilities to recog- nize a chloroplast DNA replication origin from Cblwdamonas. CHAPTER 1 RECOMBINATION BETWEEN CHLOROPLAST DNAs DOES NOT OCCUR IN CROSSES OF 0mm INTRODUCTION Recombination between chloroplast markers has been observed in the unicellular green alga, Cblmdamonas (reviewed by‘Gillham, 1978). In that organism, recombination between chloroplast DNAs (chNA) has facilitated the analysis of chloroplast gene function (Gillham, 1978; Mets and Geist, 1983). With higher plants, investigations into recom— bination have relied primarily on experiments using somatic hybrids gen- erated from protoplast fusion. Most regenerated somatic hybrid plants carry plastids from only one parent or the other, and initially no ' recombination between different plastid genomes (plastomes) was detected (Chen et 81., 1977; Douglas et a1., 1981; Maliga et a1., 1981; Flick and Evans, 1982; Schiller et 81., 1982; Fluhr et a1., 1983, 1984). The first example of chNA recombination in higher plants was recovered by selec- tion from a population of 1.9 x 105 calli resulting from cell fusions between two Aficotiana lines (Medgyesy et 81., 1985). In earlier somatic cell fusion experiments, the limited sample sizes and the lack of good selectable markers may account far the failure to detect chloroplast recombination. An attempt was made to search for chNA recombinants in sexual crosses of Oenotbera. In this plant, plastids are often transmitted at a high frequency from both parents in sexual crosses (Schotz, 1954, 1974). Thus, two different plastid types can be combined in the zygote (Meyer and Stubbe, 1974) where there should be ample opportunity for plastid fusion and plastome recombination. An earlier report (Kutzelnigg and Stubbe, 1974) indicated that plastome recombination had never been recovered in crosses of Genotbera. However, the great utility of chNA recombination as a tool for the study of higher plant chloroplast genetics compelled us to look further for such events. oeaotbera further recommends itself to this type of study for the fol- lowing reasons: 1) 200-500 seeds can be obtained from a single cross, thus providing a large offspring sample size. 2) Within each Cenotbera subsection, the plants are generally interfertile, allowing great flexibility in crossing strategies (Stubbe and Herrmann, 1982). Within the subsection Euoenotbera, Stubbe (1960) has defined five basic plas- tome types based on genetic criteria, which can be distinguished from each other also on the basis of DNA restriction fragment patterns (Gordon et ad., 1982). 3) The five plastid types have different abilities in transmission when placed in competition with other plastid types (Schétz, 1964, 1974). A "strong” transmitting plastome type in the paternal parent results in a higher percentage of progeny carrying plas- tids from both parents. 4) A wide variety of plastome mutants is avail- able (Stubbe and Herrmann, 1982). The experimental plan was to cross plants carrying different plastome types, with each type marked by a mutation that determines a white or pale green phenotype and by a distinguishable chNA restriction pattern. The progeny were analyzed for normal green sectors, which would suggest that recombination had occurred between the two mutant loci, resulting in a wild—type segregant. MATERIALS AND METHODS Plant Material. Seeds for the various Oenothera plastid mutants (Table l) were kindly provided by Prof. Wilfried Stubbe at the University of Dfisseldorf. For each of the mutants, the Roman numeral indicates the plastome type, while the Greek letter is the mutant designation. As determined by Hallier (1980) and summarized by Stubbe and Herrmann (1982), the plastome I mutants used here and the plastome III-gamma mutant have defects in photosystem I; plastome mutant II-beta has an ab- normal ribulose-l,5-bisphosphate carboxylase-oxygenase (Rubisco) activity. The II-iota and IV-alpha mutations appear to be different from some of the other mutations based on ultrastructural characteris- tics (Kutzelnigg et a1., 1975a,b); the IV-alpha mutation affects photosystem I, while the II-iota mutation is pleiotropic (R.G. Herrmann, personal communication). Crosses and Selection of Presumptive Recombinants. Since most chloroplast mutations of higher plants involve photosynthesis-related 10 defects, they would be lethal if grown in the homoplasmic condition in the greenhouse or in the field. In order to obtain sexually mature plants carrying these defective plastids, the white tissue is maintained in combination with photosynthetical1y-competent green tissue in chimeric plants. Crosses were made between such plants, having a periclinal variegation pattern and carrying one of eleven different Oenotbera plastome mutations in their L2 tissue layer (Table 1). For any given leaf, this tissue layer is continuous with the germ line of the flower occurring at the same node (Kutzelnigg and Stubbe, 1974). Thus, it is possible to recognize flowers that should have germ lines contain- ing only mutant plastids. The mutants of plastomes I, II, and III were maintained with wild-type plastome IV in the variegated plants, while mutant IV—alpha was in a plant carrying wildetype plastome I. Seeds were surface-sterilized with 20% commercial bleach (Big Chief, Sodium hyper— chlorite 5.25%), and germinated on 0.8! agar medium without growth regulators. Seedlings in the cotyledon stage were examined under a dis- secting microscope for green sectors on the cotyledons. Seedlings having green sectors were propagated through shoot culture on NT agar medium (Nagata and Takebe, 1971) with hormone concentrations described by Stubbe and Herrmann (1982). Green shoots were selected during each transfer. When adequate numbers had been propagated, roots were induced by placing the shoots on medium containing 15 ug/ml naphthalene acetic acid. The plants were then transferred to soil and grown in the greenhouse. 11 Isolation of Chloroplast DNA. Mature leaves were homogenized in 68 (w/v) sorbitol, 6 mM EDTA, 1 mM ascorbic acid, 3 mM cysteine, 0.3% (w/v) polyvinyl pyrrolidone (PVP), 0.1% (w/v) BSA, pH 7.5. The homogenate was filtered through one layer of 100 um mesh gauze and two layers of Miracloth (Calbiochem). Following differential centrifugation and a wash with the same medium lacking PVP, the chloroplasts were purified by sucrose gradient centrifugation. The chloroplasts were lysed with l mg/mfl pronase (Sigma) and 1x (w/v) sarkosyl (Sigma). The chloroplast DNA was purified by CsCl gradient centrifugation. Analysis of Chloroplast DNEs. chNAs were digested with Ham HI restric- tion endonuclease (BRL) and compared by agarose slab gel electrophoresis. RESULTS Selection of presumtive plastme recodainants. The number of seedlings examined from each cross of the first field season is shown in Table 1. Only eleven seedlings out of 7671 had small green sectors on the cotyledons and all of these variegated progeny came from the cross be- tween III-gamma (maternal) and IV-alpha (paternal). The green sectors could have arisen from recombination between the two mutant plastomes, simple complementation, or transmission of a wild-type plastome from green tissue of the variegated parents. As a control, the parental plants had been self-pollinated. The offspring of the self crosses were examined: 18 out of 379 seedlings issuing from the III-gamma self- pollination were solid green; the selfepollination of the IV—alpha 12 Table 1. Phenotypes of progeny frm crosses of plastme mutants (first season). The nuclear genotype is Oenotbera bookerl’ str. Johansen (AA), ex- cept as noted: i'albilaeta (AB) and balbirubata (AB). Mutants of plastomes I, II, and III were grown in the field in combination with wild-type plastome IV. The one mutant of plastome IV was grown in combination with wild-type plastome I. The crosses were performed as described by Stubbe and Herrmann (1982). Seeds were surface sterilized and germinated on 'sucrose-supplimented agar medium. The seedlings were scored about two weeks after germina- tion for the coloration of their cotyledons. l3 @164- Parental plastid type Number of progeny M x P White Green Variegated Total IV-alpha x I-zeta 384 0 0 384 IV-alpha x I-eta 471 0 0 471 IV-alpha x I-nu 625 0 0 625 IV-alpha x I-chi 425 0 0 425 IV-alpha x I-omega 825 0 0 825 IV-alpha x I-psia 8 0 0 8 IV-alpha x II-iotaP 294 0 0 294 IV-alpha x II-beta! 34 0 0 34 IV-alpha x III-gamma 811 0 0 811 III-gamma x I-zeta 348 0 0 348 III-gamma x I-eta 274 0 0 274 III-gamma x I-chi 855 0 0 855 III-gamma x I-nu 95 0 0 95 III-gamma x I-omega 490 0 0 490 III-gamma x I-psia 24 0 0 24 III-gamma x II-epsilon 178 0 0 178 III-gamma x II—iotaP 224 0 0 224 III-gamma x II-beta? 338 0 0 338 III-gamma x IV-alpha 957 0 11 968 l4 parent yielded 2 variegated seedlings out of 462 total. These results suggested the 11 presumptive recombinants could be due to transmission from the dark green wildrtype tissue of either parent. The reciprocal cross (IV-alpha X III-gamma) would be more likely to: yield a high frequency of progeny inheriting plastids from both parents (Chap. II), and yet no green sectors were observed. This also suggests that inadver- tent transmission of wild-type plastids from the chimeric background might have occurred. Attempts were made to establish the 11 variegated seedlings resulting from the III-gamma X IV‘alpha cross as shoot cultures. Two cell lines were lost due to contamination, and in a third the green sector did not persist. DNA was isolated from the eight remaining presumptive recom- binants for further analysis. Chloroplast DNA Analysis. Chloroplast DNA was isolated from the plas- tome mutants III-gamma and IV-alpha as well as from wild-type plastomes III and IV. Using the BamHI enzyme, their restriction patterns were compared. As shown in Figure 1A, no major DNA alteration can be detected between either mutant and its wild-type counterpart. Since chloroplast DNA from the mutant plants was less readily available, chNA from the wildrtype plastomes was used for most of the subsequent restriction pat- tern comparisons. Recombinants derived from the III-gamma X IV-alpha cross would be ex- pected to have plastome DNA with DNA fragment patterns similar to plas- 15 Figure l. CpDNA restriction fragment patterns. CpDNA was digested with BamHI and electrophoretically separated on 1.03 agarose gels. A, DNA restriction frag-ant patterns of wild- type and mutant plastomes III and IV. Lanes 1 and 2 contain wild- type and mutant chNA from plastome III of 0. grandjflora. Lanes 3 and 4 contain wild-type and mutant chNA from 0. ampbila. B, chNA restriction fragment patterns of eight presumptive recom- binants and wild-type plastomes I, III, and IV. Lanes a-h contain chNAs from the eight presumptive recombinants. 16 h 9 f IlllNobcde : an“ a» Sun” H. nun—H a” a”? g!” _ ”Mafia” 4 .5” Figure 1. 17 tome III, plastome IV; or a combination of the two. In contrast, if the green plastids came from the transmission of wild-type plastomes from the parental plants, either plastome I (maintained with IV-alpha) or plastome IV (maintained with III-gamma) should be detected. It is clear from Figure 13 that all eight presumptive recombinants have the same Bam HI digestion patterns as plastome I. Apparently, they all resulted from the inclusion of cells containing wildrtype plastids in the germ line of the paternal plant. In the second field season, crosses between five different plastid mutants were performed. A significant frequency of progeny were pure green or variegated (Table 2). Judging by the transmission patterns of wildrtype Genotberanplastids (Chapter 2 of this thesis), these green plastids also can be traced back to one of their chimeric parental plants. The pure green progeny probably obtained their wildrtype plas- tids from the maternal side, whereas, the green plastids in the variegated progeny could be from either parent, depending on the type of plastome they represent (refer to Chapter 2 of this thesis). The plants that probably had carried wildrtype plastids together with mutant plas- tids in their gens lines were indicated with asterisks in Table 2. Al- most all of the crosses involving mutant IV-alpha resulted in some variegated or green progeny due to the inclusion of wild-type plastids in the germ layer. Plastid mutant IV-alpha was maintained as a chimera with wildrtype plastome type I. Since plastome type I can transmit it- self very successfully when contributed by either parent (Chapter 2 of this thesis), it is not surprising to find the contaminated wild-type l8 Tdfle 2. Phenotypes of prog-y fru crosses of plastome mutants (second season). The variegated parental plants carried the plastome mutants indi- cated in. the maternal (M) and paternal (P) parents. The green plastids of the chimeric parents were of the plastome type IV, ex- cept for the plants which contained mutant IV-alpha. The latter carried wild-type plastome type I plastids. 19 Elego Parental Plastid type Number of progeny M x P White Green Variegated Total I-eta x II-epsilon 447 0 0 447 I-eta* x III-gamma 279 7 0 286 I-eta x IV-alpha? 449 0 3 452 I-zeta x II-epsilon 212 0 0 212 I-zeta x III-gamma l7 0 0 l7 I-zeta x IV-alphaF 35 0 l 36 II-epsilon x I-eta 176 0 0 176 II-epsilon"K x I-zeta 592 17 4 613 II-epsilon x III-gamma 10 0 0 10 II-epsilon x IV-alpha* 51 0 1 52 IV-alpha x I-eta 162 0 0 162 IV—alpha? x II-epsilon 21 16 0 37 (t) indicates the parental plants that may have carried wild-type plas— tids in addition to the indicated mutant plastids in their germ lines. 20 plastome type I in some progeny of crosses involving mutant IV—alpha. Wild-type plastome type IV was the wild-type plastid in the variegated plants carrying the rest of the plastid mutants. As plastome type IV has a very low transmission frequency when contributed from the paternal side (Chapter 2 of this thesis), the green progeny in crosses I-eta x III-gamma and II-epsilon x I-zeta are likely to be wild-type plastome type IV contributed from the maternal side. In the face of such a high background, no attempt was made to screen all of these variegated progeny as before in order to recover true recombinants from the progeny of this field season. DISCUSSION For the majority of higher plants in which plastids are transmitted only from the maternal side, the opportunity for different chloroplast DNAs to recombine is rare. Although chloroplasts with different genetic - markers can be brought together into the same cell by fusion of somatic cells, here too is little opportunity for recombination. The segregation of chloroplast markers is usually complete at the stage when a plant is regenerated, and, hence, the fusion products generally display only one of the parental chloroplast types (Chen et a1., 1977; Douglas et a1., 1981; Maliga et a1., 1981; Scowcroft and Larkin, 1981; Flick and Evans, 1982; Schiller et a1., 1982; Akada et 81., 1983). Variegated plants con- taining persistent mixed chloroplasts within individual cells were found in some fusion experiments (Fluhr et a1., 1983; Gleba et 31., 1985), yet there was no evidence for recombination between chloroplast DNAs (Fluhr et a1., 1984). If recombination does occur between chloroplast DNAs but 21 at a low frequency, the sample sizes of most somatic fusion experiments were probably too small to permit detection. Indeed, by using a large sample size (2 X 103 to 2 X 104 heterokaryons) and four selectable markers, one somatic hybrid line of Nficotjana'con- taining a recombinant plastid type was recovered (Medgyesy et a1., 1985). In this single recombinant plastome, at least six recombination sites are indicated frsm.the physical map constructed from the restric- tion patterns of the chNA. In all but one of the crosses from the first field season described here, the slower-multiplying plastids were carried by the female parent which contributes larger numbers of plastids. The faster multiplying plastids were transmitted through the pollen. Thus, in most crosses, 25- 50* of the offspring should contain chloroplasts from both parents in the cotyledon (Schfitz, 1954; Sears, 1983; Chapter II of this thesis). Despite ample opportunity for different plastomes to recombine, only 11 out of 7500 progeny contained green sectors. According to the analysis of their chNA restriction patterns, these green sectors contained only plastids derived from the wildrtype tissue of the chimeric plant. Taking the results of this experiment at face value, it would appear that recombination between chloroplasts in Genotbera is very rare, much less than 0.053. Since the sites of the mutations in these particular chloroplast mutants have not been mapped, the low frequency of chloroplast recombination may be caused by close linkage of these muta- tions on the plastome. This is unlikely, for most of the mutants have 22 distinctly different lesions, based on physiological (Hallier, 1980) and/or ultrastructural (Kutzelnigg et a1., 1975a, b) criteria. Since heteroplasmic cells are not rare in the young seedlings of Cenotbera (Schétz and Heiser, 1969), and extensive chNA exchange be- tween two plastomes has been demonstrated for Nicotiana (Medgyesy et 81., 1985), the largest barrier for the recovery of recombinant plastome must result from the low frequency of plastid fusion. Only in the zygote of Chlamydamonas has chloroplast fusion been demonstrated both micros- copically (Cavalier-Smith, 1970) and genetically (reviewed by Gilham, 1978). Among the higher plants, electron microscopic evidence indicates that plastid fusion may occur in Mflmosa‘pudfica (Esau, 1972) and .Ebsta (Vaughn, 1981). In cell fusion experiments, the use of PEG to initiate the fusion of cell membranes might have facilitated the fusion between plastid membranes as well. From a critical viewpoint, it is not clear that the experiments described here would have been able to detect recombination even if it did occur. The number of cell divisions which occurred to produce the cotyledons may not be adequate to allow the segregation of the rare recombinant chNA molecules and the plastids containing these recom- binants (Michaelis, 1967). Without a selectable plastid marker (e.g. an- tibiotic resistance), it is particularly important to allow a sufficient number of cell divisions to take place before screening for green sectors. Rather than scoring cotyledons directly, it would have been preferable to induce callus formation from the cotyledon and regenerate plants from such calli. In fact, this method has been used to recover 23 the scarce paternal plastids from the progeny of sexual crosses of Nicotiana (Medgyesy et 31., 1986). In conclusion, the apparent absence of recombination between different plastomes may be due to the limited number of cell divisions allowed before screening for such an event. On the other hand, the rarity of chNA recombination may be due to the lack of chloroplast fusion in nature. The highly conservative nature of the chloroplast genome could be one consequence of the low frequency of recombination between chloroplast DNAs (Sears, 1980, 1983). CHAPTER 2 PLASTID INHERITANCE IN OEMIW: ORGANELLE GENQIE INFLUENCES THE EXTENT OF BIPARENTAL TRANSMISSION Introduction Maternal inheritance is the rule for plastid transmission in ap- proximately two-thirds of the angiosperms thus far examined (reviewed by Kirk and TilneyhBassett 1978; Sears 1980). In these cases, the absence of plastids of paternal origin in the progeny is usually correlated with the physical exclusion of plastids from the pollen generative cell or with the degeneration of plastids during pollen maturation (Hagemann, 1979; Whatley, 1982; Connett, 1987). In contrast, some genera, such as Oenotbera land Pelargonjumg have generative cells that contain numerous plastids, and exhibit high frequencies of biparental plastid inheritance. Studies on plastid transmission in Pelargonjum have concentrated on a nuclear gene locus ib'(for plastid replication) (reviewed by Kirk and Tilnebeassett, 1978). Plastid inheritance patterns in crosses between Pelarg‘onjm cultivars are predominantly under the control of the Pr a1- leles of the female parent. The proportion of offspring inheriting plas- tids from only the maternal parent, from only the paternal parent, or from both parents, depends on the maternal genotype at this locus. A 24 25 number of possible mechanisms fer Ih-controlled plastid transmission have been suggested by Tilney-Bassett and Birky (1981). In all cultivars studied, Pelargom‘um plastids themselves also have some effects on the transmission patterns: white plastids usually have very poor transmis— sion rates in comparison with wild-type green plastids. In contrast, a consistent maternal predominance in plastid inheritance has been observed in Oenotbera, as studied through interspecific crosses. Schiitz (1954, 1968, 1974, 1975) compared the transmission abilities of wild-type green plastids from 28 species of Oenotbera by using them as the female parent in crosses in which the male parent carried one of several mutant plastids. For each cross, Schiitz reported a variegation rating, which was defined as the percentage of white tissue out of all the seedlings. Based on his extensive studies, Schiitz concluded that the variegation rating of each cross was dependent on the types of plastids contributed by both parents. Three classes of plastids in the subsection Euoenotbera were identified: strong, medium and weak, with respect to their ability to compete with the plastids contributed by the other parent (Schiitz 1968, 1974, 1975). Although a strong mater- nal bias is consistently seen, a higher frequency of biparental progeny occurs when a female carrying a ”weak” plastome type is crossed with a male carrying a "strong" plastome. Independently, Stubbe categorized the Oenotbera plastomes according to plastome-genome compatability (Stubbe, 1959). After performing inter- specific crosses and observing a range of hybrid bleaching due to 26 pdastome-genome incompatibilities, Stubbe classified haploid nuclear genomes of Cenotbera'species into three major groups: A, B, and C, which in combination make six basic types of diploid nuclear genomes (AA, AB, BB, BC, CC, AC). Plastomes were classified into five groups, namely type I through type V (Stubbe 1959), according to the compatibility of each plastome in association with the six basic nuclear types of the subgenus .Euoenothera. Subsequent work has shown that these five groups are dis- tinguishable in their chlorOplast DNA restriction patterns (Gordon et a]. 1982). According to Sch6tz’ study, the strong plastids in his class- ification correspond to types I and III, medium to type II and weak to type IV (Schétz, 1968). An effect of the Oanotbera nuclear background on plastid transmission has been noted (Hamper, 1958; Schtitz, 1974, 1975). Schiitz (1974, 1975) performed crosses in which the pairs of plastids being compared remained constant but the hybrid genome varied. Although the relative order of transmission abilities was the same for the plastome types, the variega- tion ratings changed markedly in different hybrid nuclear backgrounds. Furthermore, a plastid tended to transmit itself better in a cross in which the hybrid genome of the progeny was the same as its native nuclear background. From both the Palm-ganja! and Oenothera studies, it is clear that the effect of the nuclear genome cannot be neglected when analyzing the plastid transmission pattern. However, based on Schotz’ studies (1954, 1968, 1974, 1975), it is widely accepted that Cenotbera plastids themselves are the major determinant in the process of plastid transmission, with the nuclear genome having only a minor effect th' 27 (reviewed by Sager, 1972; Hagemann, 1975; Kirk and Tilnebeassett, 1978). In order to explore the process of plastid transmission and elucidate the control of plastid multiplication at the molecular level, a greater understanding of the processes governing plastid transmission is required. In this report, the transmission abilities of Genotberajplas- tid types I through IV are compared systematically in a constant nuclear background, to determine more clearly the differences in transmission that are attributable to the plastome type alone. The experiments described here differ from previous experiments in that: (1) all the crosses were performed in a constant nuclear background, to focus on the role of the plastome in determining transmission, (2) reciprocal crosses were performed, to allow us to exclude marker effects and more accurately assess the maternal predominance, (3) all seeds were germinated and scored in a sterilized condition on sucrose-supplemented agar medium, to extend the life time of predominantly white seedlings and to allow full expansion of their cotyledons. Material and Methods Plant material and crosses. All four plastid types (I-IV) used in our crosses are photosynthetically competent in plants of nuclear type AA (Stubbe, 1959). Crosses were performed with the plants of Genotbera hookeri str. Johansen, having a nuclear type AlAl as defined by Stubbe (1959), and carrying a representative of one of the four basic wildrtype 28 plastomes of the subgenus Enoenotbera: plastome I, native plastid of 0. bookeri str. Johansen; plastome II, 0. suaveolens Grado.; plastome III, 0. Jamarcla‘ana; plastome IV, 0. atrovjrens. Plastome V was not used because it is not viable in this nuclear background due to severe plastome-genome incompatibility (Stubbe, 1959). Seven spontaneous plas- tome mutants from different Oenotbera species (Stubbe and Herrmann 1982) were used in the test crosses. Among them, I-zeta (isolated from 0. hookerl' str. Johansen) and I-beta (isolated from 0. hooker-'1’ std.) have defects in photosystem I or cytochrome f complex, respec tively while ‘ I-eta (isolated from 0. slate) has defects in both of these two photosynthetic complexes. Mutant II-gamma (originating from 0. suaveolens Grado.) has a defective photosystem II whereas III—ga-a (originating from 0. grandjflora has a defective photosystem I. The mutations in II-epsilon (derived from 0. suaveolens Fiinfkirchen) and IV- alpha (derived from 0. a-opbjla) have not been characterized. Since the plastid mutants are chlorophyll-deficient and are not able to support photosynthesis, each mutant is maintained through the propaga- tion of variegated plants that carry wild-type plastids along with the mutant plastids. Flowers having only mutant plastids in their germ lines can be recognized by the variegation pattern of the leaf at the same node: periclinal chimeras, which should contain only mutant plastids in the L2 tissue layer of these leaves, were used (Kutzelnigg and Stubbe 1974). Thus, reciprocal crosses were made between variegated (white plastid donor) and wild-type plants. These crosses are referred to as White x Green when the f-mle parent donated the white plastids or 29 Green x White when the male parent provided the mutant plastids in the cross. Germination and scoring of the seedlings. Seeds were surface sterilized with 20% bleach and germinated on MS basal medium (Murashige and Skoog, 1962) with 0.38 sucrose and 0.8% agar. Seedlings were scored for variegation after their cotyledons were fully expanded. The frequency of biparental plastid transmission is presented with a 95% confidence in- terval estimated by assuming a normal distribution of frequencies (Steel and Torrie, 1980). When the mutant plastids were contributed by the female, the area of the green sectors was also estimated as a fraction of cotyledons for each variegated seedling. Schétz (1954, 1974, 1975) and Kemper (1958) combined the areas of cotyledons and primary leaves to determine a variegation rating; however, we have not included the first pair of leaves in our estimation. The primary leaves are derived from the apical meristem which is composed of only a small number' of cells within the embryo and hence represents a much smaller population of plastids from the zygote as compared to the cotyledons. Results Progeny from reciprocal crosses between 0. hooker-1' str. Johansen carry- ing different wildetype and mutant plastids were scored for biparental inheritance, operationally defined as the appearance of variegation, that is, green and white sectors in the same seedling. Since all the mutant plastids were maintained in periclinal variegated plants, precau- tions were taken to avoid contamination of the germ line by green plas- tids introduced by aberrant cell division in neighboring green tissues 30 (Tilnebeassett, 1986). We have found that such contaminating 'green cells introduced from a neighboring cell layer tend to give rise to egg cells with only wildrtype plastids. Flowers containing these green plas- tids will produce some pure green progeny in the otherwise White x Green crosses. Pure green progeny were detected in three out of thirtybthree White x Green crosses. Thus, for these crosses in which solid green progeny were detected, the results were examined.by compar- ing with the same crosses from a second field season. Although it is easy to recognize the unexpected solid green progeny in White x Green crosses, progeny derived from eggs with mixed plastids (Stubbe, 1957; Gleba et a1., 1985) rather than pure mutant plastids would be difficult to detect. However, when the results from two field seasons were compared, the frequency of biparental transmission from one season was within the 95% confidence interval of the other. These data have been summed and are presented in Tables 3 and 4. White x Green crosses. The plastid inheritance patterns of crosses be- tween female parents carrying one of seven different mutant plastids and male parents carrying one of four different wild-type plastids are shown in Table 3. With a constant pollen source, the frequency of transmission of the male plastid varied, depending on the plastid type in the female parent. When a wildrtype plastome type I was contributed from the male parent, the progeny always included variegated seedlings, although the percentage of progeny carrying plastids from both parents varied, rang- ing from 9.7x to 40.9%. The transmission of wildrtype plastome type III from the male parent was very similar to the transmission of plastome 31 Table 3. Biparental transmission of chloroplasts in White x Green crosses' Abundance CROSSES TRANSMISSION of paternal plastids in Female Mg1e Maternal Biparental N XBiparentalb BP progenyf I-beta I 52 36 88 40.9 1 5.2 n.m.* I-zeta I 93 10 103 9.7 1 2.9 0.30 I-eta I 86 23 109 21.1 1 3.9 0.30 II-epsilon I 88 32 120 26.7 1 4.0 0.47 II-gamma I 132 47 179 26.3 1 3.3 0.40 III-gamma I 159 44 203 21.7 1 2.9 0.42 IV-alpha I 84 56 140 40.0 1 4.1 0.56 I-beta III 66 20 86 23.3 1 4.6 0.38 I-zeta III 141 8 149 5.4 1 1.9 0.25 I-eta III 346 76 422 18.0 1 1.9 0.23 II-epsilon III 260 23 283 8.1 1 1.6 0.23 II-gamma III 286 86 372 23.1 1 2.2 0.28 III-gamma III 160 61 221 27.6 1 3.0 0.25 IV-alpha III 61 62 123 50.4 1 4.5 0.43 I-beta II 392 25 417 6.0 1 1.6 0.19 I-zeta II 101 0 101 0 - I-eta II 201 4 205 2.0 1 1.0 0.06 II-epsilon II 100 2 102 2.0 1 1.4 0.13 II-gamma II 127 5 132 3.8 1 1.7 0.16 III-gamma II 160 5 165 3.0 1 1.3 0.09 IV-alpha II 104 17 121 14.0 1 3.2 0.28 I-beta IV 264 l 265 0.4 1 0.4 0.20 I-zeta IV 164 0 164 0 - I-eta IV 372 0 372 0 -- II-epsilon IV 99 0 99 0 - II-gamma IV 296 2 298 0.7 1 0.5 0.16 III-gamma IV 192 5 197 2.5 1 1.1 0.13 IV-alpha IV 230 l 231 0.4 1 0.4 0.06 3The nuclear background of both female and male parents was 0. bookerl’ str. Johansen (nucler type AlAl). Variegated female plants, centaining only mutant plastids in the germ line were crossed with lines carrying green wild-type plastomes I, II, III, or IV. bThe biparental transmission frequencies are presented with a 95% con- fidence interval estimated by the normal approximation (Steel and Torrie, 1980). cAverage total area of tissue in the biparental progeny containing paternal plastids as a fraction of the two cotyledons. l"n.m. = not measured. tY 32 type I in the frequency of biparental plastid inheritance and it was al— ways present in some fraction of the progeny of each cross. When the male plant carried wild-type plastome type II, the frequency of see- dlings with both plastid types ranged from non-detectable to 14%. Plas- tome type IV contributed from the pollen could be detected in progeny from only four out of seven crosses, with the highest frequency of progeny with biparental plastid inheritance being 2.5%. Figure 2 shows a comparison of the transmission abilities of the four wild-type plastids with the female parent held constant. The biparental transmission frequencies can be ranked in the order I>III>II>IV in crosses involving five out of seven mutant plastids. In crosses where III-gamma and IV-alpha were contributed from the female parent, plastome type III from the male parent gave a higher frequency of biparental progeny than did plastome type I. Except for crosses in which wild—type plastome type IV was contributed from the male parent, the frequency of biparental plastid transmission and the average size of the green areas in the cotyledons' of the biparental seedlings are highly correlated (Table 3). The overall cor- relation coefficient, r=0.853, is significant at the 99% confidence level. The extent of variegation in the progeny from crosses in which male parent contributed wildrtype plastome type I or III is illustrated in Figure 3. Although the frequencies of biparental plastid inheritance are similar in these crosses (Table 3 and Figure 2), when plastome type III is carried by the pollen, smaller green sectors are seen in the 33 Figure 2. Frequency of’biparental transmission in White x Green crosses. Progeny from crosses between Cenotbera bookerj str. Johansen plants carrying one of seven different mutant plastids and plants carrying one of four wildrtype plastids were scored for biparental inheritance of plastids. The bars indiCate the biparental trans- mission frequencies from crosses grouped according to the maternal plastid types. The mutant designations are abbreviated as follows I-b for I-beta, I-z for I-zeta, I-h for I-eta, II-e for II- epsilon, II-g for II-gamma, III~g for III-gamma, IV-a for IV- alpha. 34 .BIPARENTAL TRANSMISSION: W x G CROSSES fl: [Sill 7//A 7///////////. S“““.‘““~““‘§ -~ g’I’l’l’ ’l”" s é. S‘ h‘h‘xsmu ray/"II. §§§§§ 574571.172 .11. g a“’ 2 IA fi§§§ agfié s 7" Z .5 50-4 u d J d w m m w o EMOOKE to R "-9 MATERNAL PLASTIDS l-o l—z 35 Figure 3. Abundance of tissue containing paternal plastids in variegated progeny fro-[White x Green crosses. In the crosses illustrated, either wild-type plastome type I (A) or type III (B) were used as paternal plastid. The area of green tissue was scored as a fraction of the total area of the two cotyledons. The height of the bars show the percentage of progeny exhibiting the degree of variegation, which is indicated on the X- axis. The abbreviations of the mutants are the same as in Figure 2. 36 é a P///// fl/A/A/ r/W///A/A////ifl/4lfl/flgr/ \ N / N IE III-9 "-9 22 lV-o IZZI I-h m I-z -|l-c o o. 0 V .... .. O V.‘\§\\\\AA r ‘--‘-~O 3"““““‘O ‘\\\\\\\\\\\\\\\‘§\\\\5 n’”/””””””’/”’. ‘..-‘O "““"“O .m O \‘\\\\\\\ I’ll/I’ll]. “-‘. "“""“‘O .H.H.H.M.H.HO Q‘\\\\\\\\\u {Iii/I’l’llfl’lll’ “-‘~O i...“ ¢ .1 C ..... Q 4 - .......... ’. I’ll’.’”/”””””A ‘-‘-~‘-‘O fly”””/”.””A ‘---‘.-. "““““““‘ “season.“ ONOHOhOMO.0».“0w0p0n0h0u0n0»0n.u.uo V\ \\\\\\\\\\\\\\. Ill/lllllll/ lilil/ll/l/l/l. §--‘~.-:-b l“‘.““ ‘fl‘d “enoneuoneufl ..-...-.-.-..n...-.....n.n.u.n.nO ‘\\\\\‘\‘\ ’l”’l”’/”’/”/”l’l””””’i V‘-~..---.‘~‘~.~‘~‘~. 14 13" 12" 11- 10‘ J n a q _ _ a q — J 9 a 7 5 5 4 3 2 1 o >Zm00Ml to R <1/4 <1/3 <1/2 <2/3 <3/4 '<7/a >7/a <1 /3 FRACTION 0F comm 37 variegated progeny than when plastome type I is contributed by the pollen. Green x White crosses. For most of the crosses listed in Table 3 reciprocal crosses were performed. Biparental transmission frequencies of these crosses are presented in Table 4. When wildrtype plastome I or III was contributed by the female parent, the frequency of biparental seedlings in the progeny was low. In contrast, the presence of a wild- type plastome IV in the female was associated with a very high incidence of biparental plastid transmission. Wild-type plastome II in the female parent of the corresponding crosses resulted in a wide range of biparen- tal transmission depending on the plastidrtype contributed by the male parent. These data are illustrated in Figure 4, where they are grouped according to the plastid type of the male parent. With the male parent constant, the biparental plastid transmission frequencies vary according to the plastid type of female parents. Statistical analysis of biparental transmission frequencies. The biparental plastid inheritance frequencies listed in Table 3 and 4 were transformed into arcsine and subjected to a randomized complete-block analysis (Steel and Torrie, 1980). Results of these analyses are presented in Table 5. The F ratio for all three terms: the maternal, paternal, and the interaction between these two terms, are significant. Since the nuclear background is constant throughout the experiment, the plastome type is the only variable in both maternal and paternal terms. In both sets of crosses, the variance attributable to the type of pater- nal plastome is greater than that attributable to the type of maternal 38 Table 4. Biparental transmission of chloroplasts in Green x White crossesa CROSSES TRANSMISSION Female Male Maternal Biparental N XBiparental I I-beta 167 2 169 1.2 1 0.8 I II-gamma 9 0 9 0 I III-gamma 83 7 90 7.8 1 2.8 I IV-alpha 206 0 206 0 III I-beta 246 19 265 7.2 1 1.6 III I-eta 89 20 109 18.4 1 3.7 III II-gamma 96 4 100 4.0 1 2.0 III III-gamma 219 21 240 8.8 1 1.8 III IV-alpha 224 2 226 0.9 1 0.6 II I-beta 145 22 167 13.4 1 2.4 II I-eta 7 9 16 56.3 1 12.4 II II-gamma 85 0 85 0 II III-gamma 106 49 155 31.6 1 3.7 II IV-alpha 171 2 173 1.2 1 0.8 IV I-beta 135 37 172 21.5 1 3.1 IV I-zeta 215 104 319 32.6 1 2.6 IV I—eta 80 87 167 52.1 1 3.9 IV II-gamma 173 39 212 18.4 1 2.7 IV II—epsilon 216 38 254 15.0 1 2.2 IV III-gamma 107 136 243 56.0 1 3.2 IV IV-alpha 142 0 142 0 'The nuclear background of the plants are the same as in Table l. 39 Figure 4. Frequency of biparental transmission in Green x White The bars indicate the biparental transmission frequencies from crosses grouped according to the paternal plastid types. The ab- breviations of the mutants are the same as in Figure 2. 40 BIPARENTAL TRANSMISSION: G X W CROSSES IS -IV § 22] VIII. ®§x\\ I x \ 6% \\ \\\\\ .\\\\ \\\.\. .w\\\\ \\ ..\\\\ A If / /../ I A V\\ 7/// q _ q - .... m m. m 501 >ZMOOK1 .....O N IV-o _ ll-g l-b PATERNAL PLASTIDS Ill-g 41 Table 5. Analysis of variance of biparental transmission frequencies Whitalfisen d.f. MS F sip/03m Maternal 6 156.38 7.21** Paternal 3 1216.72 56.11*** 5.1 M X P 18 21.69 5.16** Error 0 4.21 61W d.f. MS F sip/61m Maternal 3 376.94 5.62* Paternal 4 769.31 11.76** 2.9 M X P 12 65.41 12.28** Error' 0 5.30 *, P < 0.05; **, P < 0.01; ***, P < 0.001 °Theoretical variance for arcsine transformation 42 plastome in the cross. Comparisons of different :mutant plastids. Crosses were performed in which the female contributed one of three different plastome I mutants. The result was three distinguishable transmission patterns (Figure 2). The transmission abilities of the three plastids when contributed from the>female parent are in the order of I-zeta>I-eta>I4beta. When III-gamma and I-eta were contributed by the male parent, these two mutants gave a high frequency of biparental progeny (Table 4). But when contributed by the female parent, their transmission abilities were relatively low, compardale to the two plastome II mutants (Figure 2). Similarly, when mutant I-beta.was used as the maternal plastid, as high a frequency of paternal plastid transmission occurred as when mutant IV- alpha was donated by the maternal parent (Figure 2). However, when mutant I-beta was contributed from the paternal side, it showed a much higher transmission ability than did mutant IV-alpha (Table 4). Discussion Genothera is known to transmit plastids frsm both parents in crosses, but not all crosses result in a high frequency of biparental inheritance (reviewed by Kirk and Tilneybnassett, 1978). The frequency of biparental transmission can be influenced by both nuclear and plastid genomes, but the relative importance of these two components is not clear. Studies of Pelarganim have demonstrated a strong nuclear control of the plastid transmission patterns (Tilney-Bassett, 1979; Tilneyhnassett and Birky, 43 1981). The experiments presented here have demonstrated that, in Genotbera, plastid transmission depends on the types of plastome and the interactions between the two types of plastids contributed by both parents. These effects are superimposed on an intrinsic maternal predominance. In this report, the tran-ission abilities of four major plastid types of subgenus Ehoenotbara’are compared in a constant nuclear background. The feur wildrtype plastids show very different transmission abilities in the nuclear background of Genotbera bookeri str. Johansen (nuclear type AiAi). Plastome type I, the most successful plastome, was trans- mitted to some fraction of the progeny in all crosses. Plastome type III appeared to have a transmission ability very close to that of plastome type I, but plastome types II and IV showed much weaker transmission abilities when compared to the others. Hence, in this nuclear background, the relative transmission abilities of the four repre- sentative plastomes are in the order of I)III>II>IV. These results agree with Schétz’ conclusions from comparisons made in various nuclear back- grounds (Schétz, 1974). How can the great variation of biparental plastid transmission be explained? Several hypotheses fer the variation of plastid inheritance have been suggested. The simplest explanation is a variation of input: gametes with different plastid types carry different numbers of plastids. Hewever, if the biparental transmission frequency simply reflects the input numbers of plastids from both parents, one would con- 44 clude that the male gametes that produce almost no biparental progeny should contain no or very few plastids. Yet the same pollen can be crossed to a different female parent and make a significant contribution to the plastid population of the progeny (Table 3 and 4). Likewise, if the female parent is held constant (such that the plastid input from the female parent is constant), different frequencies of biparental trans- mission are observed when the plastid type of the pollen is varied. Furthermore, Schotz (1954) found that plastid numbers did not differ significantly in the egg cells of different Genotbara’species, yet dif- ferent frequencies of biparental progeny were obtained when these plants were crossed with a constant pollen source. In short, differential input of plastids in the gametes cannot be the major cause of the plastid- dependent variations in Qenatbers plastid transmission. A second hypothesis is that differential destruction of plastids and/or plastid DNA destruction occurs during fertilization, perhaps through a DNA.modification/restriction system in the zygote, as postulated by Sager and Kitchen (1975). According to this theory, critical events would occur shortly after gamete fusion. However, our observations sugb gest that the extent of plastid transmission from the male parent is af- fected not only by the events of fertilization, but also by later events. As shown in Table 3 and Figure 4, crosses in which wild-type plastome I or plastome III is contributed by the male parent produce similar frequencies of biparental progeny, suggesting that they are equally successful in their initial transmission (or survival) within the fertilized egg. However, as_shown in Figure 3, biparental progeny 45 that receive wildrtype plastome I from the male parent tend to have much larger green sectors than those that receive wild-type plastome III from the pollen. This suggests that the critical stage for the determination of the degree of maternal predominance extends beyond the fertilization event. Furthermore, an electron microscopic study of zygote development in Oenotbera did not indicate degeneration of organelles from either parent following fertilization (Meyer and Stubbe, 1974). Thus, the dif- ferential destruction hypothesis seems unlikely for this plant. Another hypothesis suggested by Sager (1972) as well as Tilnebeassett and Birky (1981) to explain genetic data of biparental plastid in- heritance is the gene conversion model. This hypothesis suggests that town-bination between different plastomes in the zygote frequently oc- curs and the subsequent DNA repair is biased through the use of the ”strong" plastome as template. Existing data do not support this theory: (1) plastid fusion is infrequent in higher plants (reviewed by Sears, 1980), and (2) even under conditions designed to select infrequent chNA recombinants, none have been recovered from (biparental progeny of Genotbera (Kutzelnigg and Stubbe, 1974; Chiu and Sears, 1985). A fourth hypothesis to explain the differential transmission of Genotberaqplastids was originally proposed by Renner (1924, 1929) and also advocated by Schétz (1954). This hypothesis postulates that dif- ferent plastids have different intrinsic rates of multiplication. Ac- cording to this hypothesis, the faster replicating plastid in mixed cells will be propagated at the expense of the slower one, due to a copy 46 number limit on the plastids in each cell (Kirk and Tilnebeassett, 1978). Consequently, the frequency of the slower multiplying plastid in subsequent cell generations will be decreased. The competitive advantage of the faster plastid will exist until the two types of plastids sort out completely. Schétz and Heiser (1969) analyzed the plastid content of heteroplastidic mesophyll cells of variegated progeny and concluded that the crosses which produced a higher variegation rating also had mixed cells with a higher frequency of plastids of paternal origin. Although this is not direct evidence for differential multiplication of plastids in mixed calls, all the genetic studies reported thus far are consistent with this hypothesis. Our data support the hypothesis that the differential transmission of Obnotberaxplastids is caused, in part, by plastid-dependent differences in multiplication. First, the analysis of variance suggests that, with the nuclear background held constant, the transmission of plastids is dependent upon the plastid types of both parents and the interaction be- tween these two types of plastome (Table 5). Secondly, the plastid types that had the highest transmission frequencies when contributed from the male parents were also the most successful when contributed from the female parent. Thirdly, with the plastid from the male parent constant, the White x Green crosses that produced the highest frequencies of biparental seedlings usually gave rise to the largest sectors of green tissue in those biparental seedlings (Table 3), which is an indication of earlier plastid segregation. The number of cell divisions required for complete sortingbout is dependent on the total number of segregation 47 units and the input ratio of one plastid type to the other in the mixed cell (Michaelis, 1967). Assuming a constant plastid input at fertilization, the dependence of biparental transmission frequency on plastome type must be achieved by the relative multiplication abilities of the two types of plastid within the mixed cells. Finally, this model can partially explain the absence of plastids from the male parent in all of the progeny of some crosses: a plastid contributed by the pollen may not be detected if it has a double disadvantage of a significantly lower input as well as’a lower multiplication efficiency compared to that of the maternal plastid. However, differences in multiplication rates alone cannot explain all the observed differences in plastid transmission. If the multiplication rates were the only determinants, then one would expect to see the same biparental transmission frequencies fer all crosses in which the green and white plastids have similar multiplication rates. In fact, crosses with the "stronger" plastids and their corresponding mutants (e.g. I- zeta x I) resulted in higher biparental transmission frequencies than equivalent crosses with ”weaker" plastids (e.g. IV-alpha x IV). These results could be explained if the initiation of multiplication of the plastid from the paternal parent is generally delayed in comparison to the plastids from the maternal side and if the period of delay depends on the plastid type (Meyer and Stubbe, 1974; Hagemann, 1976). In general, plastids from the male parent in young zygotes are smaller, and they do not contain as many starch grains as plastids derived from the female (Diers, 1963; Meyer and Stubbe, 1974). It is possible that male 48 plastids take a longer time than female plastids to accumulate factors required for multiplication. The higher variance values attributable to the paternal plastome type compared to that of maternal ones (Table 5) lend support to this hypothesis. If a lag period does in fact precede multiplication of paternal plastids in the zygote, this would explain some of the discrepancy between biparental frequencies and the extent of variegation in certain crosses. Crosses in which the mutant I-zeta came from the maternal side and wild— type plastome I or III was contributed by the male parent provide an example. In these crosses, the biparental transmission frequencies were low, but the sizes of green sectors in the biparental progeny were com- parable to those crosses with higher frequencies of biparental transmis- sion (Table 3). Evidently, the two measurements are sensitive to two re- lated but separable phenomena, and we suggest that there are two stages that affect the transmission pattern: the first stage determines whether any plastids of paternal origin will be transmitted to the progeny, and the second stage determines the extent of variegation. Early embryogenic development in Genotbera’offers support for the idea that there might be two different stages determining the transmission pattern. The first division of the zygote is asymmetrical. The much larger of the two daughter cells develops into the suspensor, and the smaller terminal cell develops into the main body of the embryo (Renner, 1915). Since the mixing of plastids can be very limited in the zygote (Meyer and Stubbe, 1974), a delay in the multiplication of paternal 49 plastids compared to that of maternal plastids before zygote division may greatly decrease the probability of their entering the terminal cell and hence lower the observed frequency of biparental progeny. In later stages of embryo development, however, the relative success of the two different types of plastids in multiplication determines the relative proportion of the two different types of plastids in mixed cells, which, in turn, determines the extent of variegation in the seedling. The dif- ferences between wildetype plastome type I and III when contributed from the paternal side, as illustrated in Figure 3, could result from dif- ferential plastid multiplication in the second stage. In Pelarg‘om’um, mutant plastids generally do not transmit as well as wild-type plastids (Kirk and Tilnebeassett, 1978). However, Schotz’ (1975) studies of Genotbera suggested that plastid mutants and their corresponding'wildrtypes have comparable transmission frequencies. In our experiments, several plastid mutants showed differences in transmis- sion compared with the wildrtype plastomes. In one case, three mutant type I plastomes showed three distinguishable transmission abilities in crosses with wildrtype plastids. The plastid mutant I-zeta is very similar to the wild-type plastome I in its transmission ability, and in fact, it was isolated from 0. bookerj str. Johansss, the same source as the wild-type plastome I representative. The mutants I-beta and I-eta have much lower competitive abilities compared with I-zeta when con- tributed by the female parent. These mutants were isolated from 0. bookerj standard (nuclear type AzAz) and 0. elata (nuclear type A1A1) respectively. Although it is possible that the mutations carried by the 50 plastomes affect their competitive abilities, the variation in transmis- sion abilities could also reflect differences inherent to the various plastids that have been grouped as plastome type I. In most cases, the relative competitive ability of a plastid does not depend on whether the plastid is contributed by the female or the male parent. Crosses involving plastid mutants III-gamma, I-beta and I—eta are exceptions to this rule. These three plastid mutants showed greater competitive abilities when contributed by the male parent than by the female parent. One explanation for this difference is that some physiological functions of the maternal plastids could be important for efficient selfemultiplication and hence, a high degree of maternal predominance. Although our experiments employed Genotbera, the concepts defined by this investigation may be pertinent to plastid transmission in higher plants in general. Even in plants such as Nicotjana and Eijobjm, con- sidered to inherit plastids in a strictly maternal fashion, trace amounts of plastid DNA from the male parents can be detected in the progeny (Medgyesy at 81., 1986; Schmitz and Kowallik, 1986). In this report, we have confirmed the significant role of the plastome in the process of plastid transmission. Some of the patterns of plastid transmission can be explained in terms of intrinsic competitive abilities of the plastids, but neither competition nor input bias can account for all of the observations. The manner in which the plastome 51 participates in this process remains to be determined, as does the role of the nuclear genome in the same process. Certainly, lower input of paternal plastids alone cannot explain the predominance of maternal transmission: events occurring after fertilization, particularly com- petitive multiplication of plastids, must also be considered. CHAPTER 3 INDIRECT APPROACHES TO CHARACTERIZE CHLOROPLAST DNA ORIGENS 0F REPLICATION Introduction Plastids are not created de novo. That is, they can originate only from preexisting plastids (reviewed by Possingham and Lawrence, 1984). The jpropagation of plastids involves two major events: chloroplast DNA replication and plastid division. Most enzymes involved in these two steps are likely to be nuclear encoded (Scott et 81., 1982). However, the plastome itself has a strong influence on plastid transmission and multiplication (Schotz, 1974; Chapter 2 of this thesis). The plastome may affect these processes through the contribution of some cofactors or via special recognition sequences, such as the origin of DNA replication (ori/rep). Chloroplast DNA origins of replication have been located in Chlamydamonas (Weddell et a]. , 1984) and Euglena (Havel-Chapuis et 31., 1982) through electron microscopic localization of DNA replication in- termediates named D-loops, where D stands for displacement. The DNA sequences of these chNA replication origins reveal some common features: they are both highly A+T rich and able to form large stem-loop structures (Wu et a1., 1986; Koller and Delius, 1982). In higher plants, 52 53 replication of chloroplast DNA has been shown to initiate from two sites 7.1 kb apart (Kolodner and Tewari, 1975). However, these replication origins have not been characterized further, presumably due to the low frequency of replication intermediates that can be observed. An alternative way to identify the origin of chloroplast DNA replication would be through a functional assay of DNA sequences, testing for those that can act as initiation sites for DNA synthesis. Since no means is available currently to transform the chloroplast, it is not yet possible to test a putative origin of replication in viva._ For this reason, heterologous systems have been used to test the function of putative DNA replication origins. Both yeast (Ohtani at 51., 1984; Overbeeke et a1., 1984) and Cblamydamonas (Rochaix et 81., 1984) have been used as heterologous systems to aid the search for a chloroplast origin of replication. Sequences that can promote autonomous DNA replication in yeast are named ARS (autonomous replication sequence) (Struhl et a1, 1979). The frequency of ARS’s found in the yeast genome coincides with the abun- dance of replication intermediates observed in the chromosome (Beach et a1., 1980; Chan and Tye, 1980) and they are postulated to represent yeast nuclear DNA replication origins. However, whether AHS’s are normal DNA replication origins in yeast chromosomes is still open to question (Walmsley at 8.1., 1984). 54 Several sequences with ARS properties have been found in the chloroplast genome by transforming yeast with EL on i plasmids that contain cloned chNA fragments. At least seven such DNA regions have been identified in the Chlamydhmonas’chloroplast genome: three have been found in .Petunja (de Haas at a]. 1986), two are known in Nficotjans, and at least one ex- ists in Chlaredla (Yamada et 81., 1986). These chloroplast DNA sequences all contain the 11 bp consensus sequence common to the yeast ARS. Another heterologous system has been used by Rochaix at a]. (1984), who searched for sequences that can promote DNA replication in the nucleus of able-Jonson”. ARC (autonomous replication in Cblamydamanas) activity has been found in four fragments of Cblamrdamonas chloroplast DNA (Rochaix, 1984). Two partially conserved elements of 19 and 12 bp are found in these four ARC containing sequences. None of these ARS and ARC sequences found in chloroplast genomes ac- tually coincide with the observed initiation sites of chloroplast DNA synthesis recognized by D-loop mapping, although some are nearby (Vallet and Rochaix, 1985; Yamada et a1., 1986). The actual role of these ARS or ARC sequences in vivo is still unknown. Nevertheless, in the Chlamydhsonas in vitro»chNA replication system, a fragment containing ABS and ARC together with one of the chNA replication origins (ariA) served as a better template for DNA synthesis than ariA alone (Wu et a1., 1986). Also, a maize chNA fragment that incorporated the highest amount of nucleotides under the direction of partially purified pea chloroplast DNA polymerase also contained ARS and ARC consensus 55 sequences (Gold at 81., 1987). If autonomous replication sequences in yeast (ARS) or in Chlamydomonas (ARC) do enhance the binding of DNA replication enzymes, their presence near the chNA replication origin may facilitate the replication of the plastid genome. To test this hypothesis, three ARC sequences from the CZIamydbmonas chloroplast genome were used to probe for homologous sequences in three types of Genotberanplastomes that showed differential transmission abilities (Schotz, 1974; Chapter 2 of this thesis). Despite the failure of yeast and Chlamydhmonas to recognize the ori/rep of the chloroplast, there are examples in which the true ori/rep could be recognized by a different organism. Heterologous systems have been successfully used to characterize essential DNA sequences within the bacterial chromosomal DNA replication origin (oriC). Zyskind 8t 81. (1983) identified the replication origin of the marine bacterium PTbrio harveyi through functional analysis of DNA replication in AL codj. Com- parison of ijn'o harveyz' with five moderately related bacteria from the family of Enterobacteriaceae revealed clusters of conserved sequences composing the replication origin. Instead of using eukaryotic transformation systems such as yeast or 0h18aydomonas'to test the function of chNA ori/rep, a prokaryotic sys- tem is a more reasonable choice, since chloroplasts are almost certainly of prokaryotic origin. Cyanobacteria, with their presumed evolutionary relationship to chloroplasts (reviewed by Whatley and Whatley, 1981; 56 Doolittle, 1982), would seem to be the best candidate for an in vivo heterologous system to functionally analyze a chloroplast DNA replica- tion origin. In the experiments described here, plasmids carrying a Chlasydhmonas chloroplast DNA replication origin (arid) were introduced into three different strains of cyanobacteria either by triparental mating or by transformation to test if this chloroplast DNA origin of replication can support autonomous replication in these heterologous systems. Material and Methods Isolation of chloroplast DNA froml Chlamwdomonas. A strain of Chlamwdolonas reinbardii cc-1615 (from the culture collection at Duke University), lacking a cell wall (cw-15 mutation) was used as a source of chloroplast DNA. Chloroplast DNA was isolated according to Rochaix (1984) with some modifications. One to four liters of ChJamydomonaS'cell culture were grown in Tris-Acetate-Phosphate (TAP) medium (Carmen and Levine, 1965). Stationary cultures were harvested by centrifugation at 3,000xg for 5 minutes. The pellet was resuspended in NET (15 mM NaCl, 100 d4 .TA, 50 IN Tris, pH 9.0) buffer at a final density of 1 x 109 cells/m1. Predigested pronase (2 hours at 37°C, 10 min. at 80°C) was added to a final concentration of 0.6 mg/ml. The solution was mixed in the cold for 10 minutes and 10% SDS was added to a final concentration of 1.2:. The solution was incubated at 50°C. Additional pronase (half the amount of the first addition) was added after one hour. After 2 to 2 1/2 hours of incubation, the lysate was cooled on ice and extracted with 57 phenol/chloroform. After ethanol precipitation of the nucleic acid, the high molecular weight nucleic acid was spooled on a glass rod and im- mersed in cold 70% ethanol for 10 minutes. After drying the nucleic acid on the glass rod in air, the total nucleic acid was resuspended in TE (10 11M Tris, 1 1m EDTA) buffer and treated with RNase (50 ug/ml, 37°C for one hour). The solution was extracted with phenol/chloroform and the DNA was spooled as before. The concentration of DNA was adjusted by measuring optical density at 260nm (A250) and diluting the DNA by adding 3-5 ml TE buffer for every 10-15 Aaso units. Bisbenzimide was added to a final concentration of 100 ug/ml. CsCl was added to the solution and the concentration was adjusted to a refractive index of 1.3980. After an overnight centrifugation at 45,000 rpm in a vT165 rotor, four bands could be resolved. The species of DNA bands from top to the bottom are chloroplast, mitochondrial, ribosomal and nuclear DNA, respectively (Rochaix, 1984). DNA blot hybridization. Chloroplast DNA from three major types of Genotbera ‘plastids were digested with the restriction endonucleases BamHl and EcoRl. The fragments were separated on an agarose gel and transferred onto nitrocellulose paper. Plasmids carrying ARC sequences (pCAl, pCA2, pCA3, pCA4) were obtained from Dr. J .-D. Rochaix. Each ARC- containing fragment (ARCl, 153 bp; ARCZ, 414 bp; ARC3b, 257bp) was cut out from the vector, pJD2, after digestion with Hind III and SalI. These fragments were separated from the vector on an agarose gel and were then removed from the gel by electroelution. Each ARC-containing fragment was ethanol-precipitated, redissolved and then labeled with 32P—CTP by nick 58 translation (Maniatis et 81., 1982). The hybridization and washing pro- cedures were performed in low stringency (Maniatis et 81., 1982) such that the three ARCs can cross hybridize with each other. Cloning of the chloroplast DNA origin of replication fru mm. Two plasmid vectors, pRL178 and pRL424 (Elhai and Wolk, submitted) were used in this study. These vectors are based on the replicon of pBR322. The common features of these plasmid vectors include a symmetrical polylinker for cloning and the kanamycin resistance gene from Tn5 as a selectable marker. A 5.6 kb EcoRI (R-l3) fragment that contains a Ch1amydomonas reinbardii chloroplast DNA origin of replication (Weddell et a1., 1984) was isolated through agarose gel electrophoresis as described previously and was cloned between the EcoRI sites of cloning vector pRL178 and pRL424 (Figure 5). This 5.6 kb fragment was also sub- cloned as 3.6 kb and 2.0 kb fragments by using the single BamHI site in- side of the R-13 fragment and the two BamHI sites on the polylinker. Bacterial strains and growth conditions. The unicellular cyanobacterium Synecbococcus 122 and the filamentous cyanobacterium Anabaena strain M131 were used for the conjugation experiments. These strains were obtained from.C.P. Wolk (MSU). synecocystis'PCC 6803 was obtained from L. McIn- tosh. (MSU) and was used for transformation. A11 cyanobacterial strains were grown in liquid in either BG-ll (Rippka et 81., 1979) or eight-fold diluted A.& A. (Hu et 81., 1981) medium supplemented with 10 mM NaNOa. These cultures were maintained under constant light (1500 1x), shaken (100 rpm) at 30°C. 59 Figure 5. Maps of the plasmids carrying 6'. reinbardii chNA oriA. The 5.6 kb EcoRI (R13) fragment of CL reinbsroii chNA, which con- tains one initiation site of chNA replication (oriA) was cloned into the vectors pRLl78 (A) and pRL424 (D) in both possible orien- tations (clone names indicated at right). The thick arrow inside of R13 indicates the position and the direction of the initiation of DNA synthesis at oriA. The regions between the two HinDIII sites of both vectors contain multiple cloning sites (MOS). There are two SmaI (AvaI) sites in the MOS of pRLl78 but none in the MOS of pRL424. The thin double-sided arrows indicate the inverted repeat regions within the vectors. The bom' (basis of mobility) region is required in cis for conjugal mobilization. The opt (neomycin phosphotransferase) gene confers resistance to kanamycin, neomycin, and some other aminoglycosides. ori repre- sents the replication origin of the plastid vector. The directions of transcription and DNA replication are indicated by half-arrows. 60 X r x 14.3ch I ' poms-s l c\" ,s6 C\3 I 1 I 1 ' pCR13-9 Qx 8t} x x 9.6“ :pCR13-II prlpCR13-I 6"" PhI 61 Conjugation of plasmids into cyanobacteria. Conjugation of plasmids. into cyanobacteria was performed according to Walk at 81. (1984). Over- night cultures of £2 801i strains containing conjugal plasmid (RP4 [Km'Tc'Apr]) or helper plasmid (pDS4101 [Apr]) with or without test plasmids (containing chloroplast DNA fragments [Km']) were diluted 40- fold and regrown for 4 hours. Each strain was harvested by spinning 1.5 ml of the cultures in microfuge tubes for one minute. The cell pellet was resuspended in 1.5 m1 of LB medium. 0.75 ml of EC coJi cells con- taining the conjugative plasmid (RP4) and the same volume of cells con- taining the test plasmid were mixed and centrifuged together. 60 ul of LB was added to the pellet. At the same time, cyanobacterial cells were prepared. 45 m1 of a growing culture was centrifuged at 2000 rpm for 10 minutes. The cell pellet was resuspended in 1 m1 of BG-ll liquid medium. A series of dilutions from 1 to 107-fold was made from this cell suspension. 5 ul of each EL 8011 mating mixture was added separately to 5 ul of cyanobacterial cells from each dilution. 2 ul of each EicoIi/cyanobacteria mixture was spotted onto a nitrocellulose filter placed on a BG-ll agar plate. These plates were incubated at 30°C with continuous illumination. After 24 hours, the filter was transferred to a BG-ll agar plate containing 2 ug/ml of kanamycin. In the case of Anabaena M131, 10 ug/ml of neomycin was used. Since neomycin phos- photransferase confers resistance to either kanamycin or neomycin, the choice of antibiotics is determined by the sensitivities of the organisms. 62 Transformation of Synechocystis 6803. The transformation of synecboqms- tis 'was performed according to a modified procedure of Grigorieva and Shestakov (1982). A 30 ml culture of Smechocystis 6803 was started in BG-ll liquid medium at an optical density of A730 = 0.05-0.15. The cul- ture was grown to A730 = 0.37. The cells were harvested by centrifuga- tion at 6,000 rpm at room temperature for 10 minutes. The pellet was resuspended in fresh BG-ll at an optical density (A730) of 2.5, which corresponds to a cell concentration of about 2 X 103/ml. 0.3 ml of cell suspension was mixed.with 1-2 ug of the plasmid DNA to be tested. The transfermation mixture was incubated at 34°C for 2-4 hours with oc- casional shaking. 100 ul of the transformation mixture was plated onto nitrocellulose filters placed on BG-ll agar plates which were then in- cubated at 34°C for 18-20 hours. The filters were transferred to BG-ll plates with either 5 or 10 ug/ml of kanamycin. Isolation of total DNA from Smecbocystis PCC 6803. DNA from synecbocystis' was isolated according to a base lysis procedure (Kuhlemeier, 1981). A 500 ml culture of synecbocystis 6803 was grown to stationary phase with aeration. The cells were spun down and the pellet was washed once with 10 ml of NE solution (0.12M Na01, 0.05M EDTA), and again with 10 ml of lysis buffer (23 H4 Tris, pH 8.0; 10 1|“ EDTA; 50 #4 glucose). The cell pellet was resuspended in 1 m1 lysis buffer plus 0.5 ml lysozyme (10 mg/ml). The suspension was incubated at 37°C for one hour. 0.5 m1 of fresh 10% SDS was added at the end of the first hour and the incubation was continued for another hour. 0.25 ml 5M NaCl was added to the cell lysate and the suspension was mixed gently and left over— 63 night at 4°C. The total nucleic acids were spun down at 10,000xg for 30 minutes. The pellet was dissolved in TE buffer and was then extracted with phenol/chloroform. Total nucleic acids were precipitated with ethanol and resuspended in TE. Results Detection of sequences homologous to ARCs in abnotberarchNA. The fragments of cmmdasonas chloroplast DNA that support autonomous replication in the nucleus of this alga were used as probes to see if homologous sequences exist in Genotbera'chloroplast DNA. Two probes, one carrying ARCl and the other ARCZ, both hybridized to similar fragments of Genotbera'chNA that had been digested with EcoRI (Figs. 6B and 7B). The ARCl hybridized to three EcoRI fragments of plastome I and III DNA with approximate sizes of 4.8, 3.9 and 2.5 respectively (Fig. 60). The same probe also had hybridized weakly to a 1.6 kb DNA fragment from plastome III. Only the 4.8 kb fragment of plastome IV could be detected when probed with the ARCl-containing fragment. The ARCZ probe hybridized with two fragments of plastome I (4.8 kb and 3.9 kb), three fragments of plastome III (4.8 kb, 3.9kb and 1.6kb), and two fragments of plastome IV (4.8 kb and 1.6 kb). ARC3b is located inside the gene encoding the chloroplast 23s rHNA. A fragment containing this region hybridizes strongly with the 4.8 kb fragment (Fig. 6B). This 4.8 kb EcoRI fragment recognized by all three probes is located inside of the inverted repeat region of the Genotbera chloroplast genome (Gordon et 81. , 1981). The 1.6 kb EcoRI fragment which hybridized to the A1101 and A1102 probes is only present in plastome types 111 and IV (Fig. 6B). Figure 6. Hybridization of A1101 element to Genotbera CpDNA. .At Agarose gel electrophoretic patterns of CL reinbaroij (C), 0enotbera plastome type I, III, and IV chNA digested with EcoRI. B. Autoradiogr-s of Southern hybridizations of the DNAs described above probed with the 153 bp fragment containing ARCl, which is located inside of EcoRI fragment 13 (R13) of Oh reinbaroii chNA. frhe numbers indicate the sizes (in kb) of the bands hybridized by the probe. Figure 6. 66 Figure 7. Hybridizations of ARC2 and AHC3b elements to Oenothera chNA. Autoradiograms of Southern hybridizations of the chNA described in Figure 6 with the 32P-labelled fragments carrying ARC3b (A) and ARC2 (B). ARC3b and ARCZ are located inside of EcoRI fragments 24 and 18, respectively, of CL reinbardii chNA. The sizes of the bands hybridized with the probes are indicated in kb. 67 Figure 7. 68 Testing autonuous replication activity of oriA in cyanobacteria Conjugation of an oriA—containing'plasmid.into cyanobacteria. The 5.6 kb R713 fragment that contains the oriA of Chlamydomonas’chNA was first cloned into the EcoRI site of the vector pRLl78 and was then transferred into cyanobacteria by conjugation. The two clones containing the R-l3 fragment in the two possible orientations are designated as pCR13-5 and pCHl3-9 (Fig. 58). A broad-host plasmid pRL153 (Hm?) was used as a positive control for the triparental mating. There is phenotypic evidence on replication of this plasmid in Anabaena M131, Synechococcus R2, and several other cyanobac— teria (T. Thiel, personal communication). Spots containing the highest concentration of cyanobacteria always showed some growth despite the presence of kanamycin or neomycin. In one conjugation experiment, using Synechococcus R2, cell growth could be observed with pCR13-9 or pRL153 as the test plasmid, even at the level of 107-fold dilution (14 colonies from each strain). This represents a frequency of of 7_x 101° resistant colonies/ml of conjugation mixture. pCRlB-5 gave a slightly greater num- ber of resistant colonies (loll/ml) than the other two plastids, but the size of the colonies was smaller. However, when several individual colonies were streaked onto fresh BG-ll agar plates with 2 ug/ml of kanamycin, no kanamycin-resistant colonies were recovered. In the control, cells receiving no test plasmid did not produce single colonies on the conjugation plate. Only one or two colonies of Anabaena M131 were observed on the spot of ten fold dilution after receiving pCR-5 or pCR-9. The positive control, 69 pRL153 produced colonies at a lOOO-fold dilution of Anabaena. None of the colonies receiving pCR-5 of pCR79 could be recovered after streaking on fresh plate containing 5mg/ml of neomycin. Conceivably, the high concentration of kanamycin resistant £1 codi sur- rounding the cyanobacteria might protect Km-sensitive cyanobacteria against kanamycin leading to apparent Km—resistant colonies on the con- jugation plate, which would be unstable upon transfer. If so, early elimination of £1 coli following conjugation should reduce the occur- rence of these false positives. In order to eliminate the AL colj shortly after conjugation, lysogens were constructed using a strain of phage lambda carrying the temperature sensitive repressor, cIss7. Shift- ing to the non-permissive temperature should kill unwanted E. coli by inducing the phage to enter its lytic cycle. To this end, all £1 co1i strains involved in the earlier conjugation experiment were infected with phage lambda c1057 at 30°C. Temperature-sensitive lysogens were verified by selecting the EL coJi colonies that could grow at 30°C but not at 37°C Following the conjugation procedure described earlier, the temperature was raised to induce the lytic cycle of phage lambda. Upon raising the temperature to 42°C, all strains of synecbococcus R2 grew much faster on the conjugation plate. However, no differences could be observed among strains receiving pCR13-5 and pCR13-9, or the negative control pRLl78. All these strains gave rise to several colonies at 103-fold dilution of A. nidulans R2 cells. 70 Restriction of incoming DNA by endogenous nucleases of the cyanobacteria is one possible reason for the failure of the Ch1amydhlonas’ chloroplast oriA containing plasmid to survive after transfer (Walk at 81., 1984). To partially overcome the possible restriction problem, the R-13 frag- ment was transferred into a vector, pRL424, that has fewer sites that can be recognized by AvaI; an isoschizomer of which is found in Anabaena M131. The restriction specificities of enzymes from Synechococcus R2 are unknown. Clones of R—l3 in pRL424 in both possible orientations were designated as pCRl3-I and pCRl3-II (Fig. 1). These two plasmids and their subclones (pCRl3-36.4, pCRl3-36.6, pCR13-20.l, pCRl3-20.2) were sent into Synechococcus R2 and Anabaena strain M131 by triparental mating. As a control, pRL424 alone was also sent into the same strains. In each case, growth of the cyanobacteria on antibiotic-containing media occurred only at spots where high concentration of the recipient cyanobacterial cells were present. All of these clones supported the growth of cyanobacteria on the conjugation plate to a similar extent. The difference in efficiency between these clones and the vector alone was less than ten fold. Several colonies were transferred to fresh BG-ll agar plates. None of these clones gave rise to single colonies in the presence of antibiotic. Transformation of cyanobacteria.with oriA—containing'plasmids. All the clones containing portions of the R-l3 fragment of Chlamydhmonas rein- baroii chloroplast DNA were used in an attempt to transform the unicel- lular cyanobacterium synecbocystis’strain P006803. A plasmid, pKW1189 71 (Km'), which can integrate into the synecbocystis 6803 chromosome through homologous recombination (McIntosh, personal communication) was used as a positive control. Results from the first of these experiments are listed in Table 6. pRLl78 carrying the 5.6 kb R-l3 fragment from the Chlamydomonas'chNA in both orientations (pCRl3-5, pCR13-9) did not transform synecbocystis’ 6803 with a frequency higher than the vector alone, whereas many colonies were formed from the positive control (pKW1189). The same fragment cloned in pRL424 (pCRl3-I) gave a slightly higher frequency of colonies than did the vector alone. When pCR13—II, which contains the insert as pCR13-I but in different orientation, was used, the transformation frequency was as high as with the control plastid. However, it- was noticed later that this pCRl3-II plasmid preparation was mixed with the positive control plastid in a ratio of 5:1. DNA isolated from the wild type strain of synecbocystis*6803 and the kanamycin resistant colonies receiving the mixture of pCRl3-II and the positive control, pKW1189, were compared by southern hybridization using both plasmids as probes. The hybridization indicated the presence of only pKW1189 integrated into the chromosomal DNA of Synechocystis 6803 through homologous recombination but there was no indication of pCRl3-II in the cell (data not shown). In the subsequent transformation experiments, neither pCRl3-I and pCRl3- II isolated from different plasmid preparations nor their subclones (3.6 kb and 2.0 kb) showed any transformation frequencies higher than the 72 Table 6. Transformation of Smecboczgtis PCC 6803 Plasmid Number of golopies tested 5ug/ml Km lOug/ml Km pCRl3-5 0 0 pCRl3-9 7 l pRLl78 7 5 pRL153 3 O pCRl3-I 37 42 pCRl3-II* >1000 >1000 pRL424 10 9 pKW1189 >1000 >1000 I"Contained 158 of pKW1189 73 background of the vector alone. Discussion Ch1amydomonasuARC sequences themselves are not the origin of chNA replication (Rochaix et 81., 1985), however, the fact that they can sup- port autonomous replication in the nuclei this green alga (Rochaix et 81., 1984) and the localization of one of the ARC sequences (ARCl) near the oriA of ChJamydhmonas'chloroplast DNA suggest that ARCs may play a role in facilitating the binding of proteins that participate in the process of DNA replication (Wu et 81., 1986). ARCl was used to probe maize chNA but no hybridization signal was detected under relatively permissive conditions (Gold at 81., 1987). However, sequence analysis did show a 19 bp consensus sequence in the presumed maize chNA replica- tion origin. In this report, fragments carrying three ARC sequences were used to probe for the homologous sequences of Oenothera chloroplast DNA. Under non-stringent hybridization conditions, these three fragments can cross hybridize among themselves, and furthermore, one EcoRI fragment (4.8 kb) within the inverted repeat region of the Oenothera chNA was recognized by all three fragments (Fig. 6 and 7). Two other EcoRI fragments (3.9 kb and 1.6 kb) of plastome type III were recognized by the probe containing ARCZ. The same probe also lit up the 3.9 kb EcoRI fragment of plastome type I and the 1.6 kb fragment of plastome type IV. Whether or not these differences in the ARC hybridization signals among plastome types corre- late with the differential transmission abilities of plastids (Chapter 2 74 of this thesis) still requires further investigation. The relationship (between the appearance of ARS/ARC homologous sequences and the ef- ficiency of chloroplast DNA replication cannot be drawn until more ori/rep sequences of chloroplast DNA from higher plants are charac- terized and a good functional assay for chNA replication becomes available. Neither yeast nor CMamydamanas can utilize the observed chNA ori/rep ‘from Chlamydomonas (Vallet and Rochaix, 1985; Yamada et 81., 1986). It is understandable why these system did not recognize the chNA ori/rep: the line of descent leading to eukaryotic nuclei is very different from that leading to chloroplasts (Fox at 81., 1980). On the other hand, chloroplast and cyanobacteria are much more closely related (reviewed by Gray and Doolittle, 1982; Woese, 1987). Thus, an attempt was made to use an evolutionally more related system to test for the ori/rep function of chNA. Three cyanobacterial strains, one filamentous and two unicellular, were used in order to cover a wide evolutionary spectrum within the range of existing cyanobacteria. Plasmids containing one of the Chlamydhmonas'chNA replication origins were introduced into cyanobacteria either through triparental mating or transformation. None of the tested oriA containing plasmids gave stable antibiotic resistant colonies. In the case of synecbococcus’RZ, both the vector alone and the vector with an inserted oriA fragment supported the growth of the bacteria on the conjugation plate to a similar extent. However, these colonies did not survive transfer to fresh antibiotic- 75 containing plates. When synecbocysth' 6803 was transformed with a plasmid carrying homologous sequences of its chromosome (pKW1189), the transformation frequency was very high, due to the frequent stable integration through homologous recombination. However, when the oriA containing plasmids were used to transform the same strain, the frequencies were very low (Table 6). In most experiments, the transformation frequency was not higher than the vector alone. It seems that ChJamydhsonas chloroplast oriA is unable to support the independent replication and/or maintenance of plasmids in all three strains of cyanobacteria tested. For a number of reasons, chloroplasts of rhodophytans (red algae) and crytophytans are thought to be the direct descendants of cyanobacteria (Whatley and Whatley, 1981; Gray and Doolittle, 1982). However, whether the chloroplasts of green algae and higher plants arose directly from the cyanobacteria or from some other relative is still unknown (Woese, 1987). As far as the evolutionary distance is concerned, the chloroplast ori/rep of red algae might be expected to have a better chance to func- tion in blue-green bacterial system. Other factors besides evolutionary distance must also be considered. The stable inheritance of a plasmid in a growing bacterial cell involves two separate processes: replication and partitioning (Nordstrom, 1985). Even if the cyanobacterial cells can recognize the green algal chloroplast origin of replication, the proper partitioning of foreign plasmids is 76 equally important for the maintenance of such plasmids. The chloroplast ori/rep sequences alone may not be sufficient for the proper segregation of the DNA molecules. In the experiments described here, the inability to maintain stable antibiotic-resistant colonies could be explained if plasmids carrying the chloroplast ori/rep were unable to segregate properly following cell division. CONCLUSION The multiplication and development of chloroplasts require coordinated expresion of nuclear and chloroplast genes. The study of nuclear-plastid interactions in higher plants is limited because no reliable way cur- rently is available to manipulate higher plant chloroplasts. DNA recombination is a basic tool in genetics. Recombination between chloroplast genomes has never been recovered in sexual crosses of higher plants, since plastids of’most angiosperms are maternally inherited. The first chapter of this thesis described experiments designed to test the likelihood of plastome recombination in Genotbera, a genus in which frequent transmission of paternal plastids can occur. Crosses between Oenothera plants carrying only mutant plastids in their germ lines were performed and the progeny were screened for a wild-type phenotype. In the first field season, only one out of twenty crosses between different plastome mutants yielded progeny with green sectors. However, analysis of restriction enzyme digestion patterns revealed that these green sec- tors were all derived from the wildrtype tissue of the paternal plant. In the second field season, more green plastids originating from wild- type tissue were observed but no true recombinants. Thus, recombination between chloroplast DNAs of Oenothera was not detected in these experiments. Selectable genetic markers and a higher number of cell 77 78 division cycles may be required to recover infrequent recombinants. Both nuclear and plastid genomes can influence the extent of biparental plastid transmission in canotbera. However, the degree of control by each genetic compartment has never been examined separately. In Chapter 2 of this thesis, the transmission abilities of four out of the five major plastome types of Oenothera (I-V) were analyzed in a constant nuclear background by assessing both the frequency of biparental in- heritance and the extent of variegation in the progeny. Reciprocal crosses were performed with four wild-type plastids and seven white plastid mutants. The frequency of biparental plastid transmission ranged from 0 to 56* depending on the plastid types involved in the crosses. The transmission abilities of the four representative wild-type plastids appear to be in the order of I)III>II>IV in the nuclear background of 0. bookeri str. Johansen. In general, variegated seedlings from crosses which produced a higher frequency of biparental plastid transmission also had larger sectors containing plastids of paternal origin (r = 0.853). Although the transmission abilities of most Oenothera plastid mutants are comparable to the wild-type plastids, three mutant plastids derived from species having different type I plastids showed three dis- tinguishable transmission patterns. This study confirms the significant role of the plastome in the process of plastid transmission and possibly in plastid multiplication. However, the hypothesis that the different plastid types multiply at different rates can only partially explain these results. According to 79 this investigation, the time of onset of plastid division after zygote formation may also be plastome dependent. It is known that no chloroplast gene product is required in the maintenance of the normal plastid DNA level. Hence, the observed dependence of plastid transmis- sion on plastome type may be caused by a special chNA structure, such as the origin of replication. Current investigations are directed towards an understanding of differential plastid transmission at the molecular level. These involve the characterization of initiation sites of chNA synthesis in different plastome types. Also, the importance of nuclear-plastid interactions in the process of plastid transmission is being further addressed by measuring the transmission efficiencies of plastids in different nuclear backgrounds including their native ones. The chNA origin of replication is the most likely region where the con- .trol of plastid multiplication might take place. Three DNA fragments that can support autonomous replication in Ch1amydomonas‘were used to probe for homologous sequences in Oenotherarchloroplast DNA. Some dif— ferences in the hybridization signals could be observed among the three plastome types examined. However, since the function of ARC elements is still unknown and no Oenothera chNA replication origin has been mapped, the significance of the observed DNA hybridization differences relative to plastid multiplication cannot be assessed at this time. Since there is still no method available to directly transform higher plant chloroplasts, a functional analysis of the chNA origin of replication in its natural enviroment cannot be achieved. In order to 80 look for a heterologous system to functionally analyze the chNA replication origin, 51 con plasmids containing a Chlamydomonas chNA origin of replication (oriA) were transferred by conjugation into three strains of cyanobacteria. None of these plasmids carrying oriA could be propagated and/or maintained in any of the cyanobacterial strains tested. It is possible that the evolutionary distance between blue-green bacteria and chloroplasts is still too great for the chNA replication origin to function. Alternatively, the oriA region of Chlamydomonas may not have carried enough information for stable maintainance of the replicon. LIST OF REFERENCES LIST OF REFERENCES Akada, S.; Hirai, A.; Uchimiya, H. (1983). Studies on mode of separation of chloroplast genomes in parasexual hybrid calli. I. Fraction I protein composition in unseparated hybrid callus. Plant Sci. Lett. 31:223-230 Beach, D.; Piper, M.; Shall, S. (1980). Isolation of chromosomal origins of replication in yeast. Nature 284:185-187. Boffey, S. A. (1985). The chloroplast division cycle and its relation- ship to the cell division cycle. In: The Cell Division Cycle in Plants, Society for Experimental Biology Suinar Series Vol. 26 (Bryant, J. A.; Francis, D.; eds.). Cambridge University Press, Cambridge. pp.233-246. Burgess, D. G.; Taylor, W. C. (1987). Chloroplast photooxidation affects the accumulation of cytosolic mRNAs encoding chloroplast proteins in maize. Plants 170:520-527. Borner, T.; Sears, B. B. (1986). Plastome mutants. Plant Mol. Biol. Reporter 4:69-92. Cavalier-Smith, T. (1970). Electron microscopic evidence for chloroplast fusion in zygotes of Ch1amyoomonas reiobaroi. Nature 228:333-335. Chan, 0. S. M.; Tye, B. K. (1980). Autonomously replicating sequences in saccharamyces cerevisiaet Proc. Natl. Acad. Sci. USA 77:6329—5333. Chen, K.; Wildman, S. G.; Smith, H. H. (1977). Chloroplast DNA distribur tion in parasexual hybrids as shown by polypeptide composition of frac- tion I protein. Proc. Natl. Acad. Sci. USA 74:5109-5112. Chiu, W.-L.; Sears, B. B. (1986). Recombination between chloroplast DNAs does not occur in sexual crosses of Oenothera. Mol. Gen. Genet. 198:525- 528. Cleland, R. E. (1972). Oenothera, Cytogenetics, and Evolution. Academic Press Inc., London. pp.370. Connett, M. B. (1987). Mechanisms of maternal inheritance of plastids and mitochondria: developmental and ultrastructural evidence. Plant Mol. Biol. Reporter 4:193-205. De Haas, J. M.; Boot, K. J. M.; Haring, M. A.; Kool, A. J.; Nijkamp, H. J. J. (1986). A Petunia hybridh’chloroplast DNA region, close to one of the inverted repeats, shows sequence homology with the Eugieoa graciJis 81 82 chloroplast DNA region that carries the putative replication origin. Mol. Gen. Genet. 202:48—54. Diers, L. (1963). Elektronenmikroskopische Beobachtungen an der gener— ativen Zellveon Oenothera bookerj Torr. et Gray. 2. Naturforschg. 18b:562-566. Douglas, G. 0.; Wetter, L. R.; Keller, W. A.; Setterfield, G. (1981). Somatic hybridization between Aficotians rustica’and Nicotians tabacum. IV. Analysis of nuclear and chloroplast genome expression in somatic hybrids. Can. J. Bot. 59:1509-1513. Elhai, J. and Walk, C. P. (1987) A versatile class of positive selection vectors based on the nonviability of palindrome-containing plasmids that allows cloning into long polylinkers. Gene in press. Esau, K. (1972). Apparent temporary chloroplast fusion in leaf cells of Mimosa pudica. Z. Pflanzenphys. 67: 244-254. Flick, 0. E.; Evans, D. A. (1982). Evaluation of cytoplasmic segregation in somatic hybrids of Nicotiana: tentoxin sensitivity. J. Hered. 73:264- 266. Fluhr, R.; Aviv, D.; Edelman, M.; Galun, E. (1983). Cybrids containing mixed and sorted out chloroplasts following interspecific somatic fusion in Nicotiana. Theor. Appl. Genet. 65:289—294. Fluhr, R.; Aviv, D.; Galun, E.; Edelman, M. (1984). Generation of heteroplastidic Nicotians cybrids by protoplast fusion: analysis for plastid recombinant types. Theor. Appl. Genet. 67:491-497. Fox, G. E.; Stackebrandt, E.; Hespell, R. B.; Gibson, J.; Maniloff, J.; Dyer, T. A.; Wolfe, R. S.; Balch, W. E.; Tanner, R.; Magrum, L.; Zablen, L. B.; Blakemore, R.; Gupta, R.; Bonen, L.; Lewis, B. J.; Stahl, D. A.; Luehrsen, K. R.; Chen, K. N.; Woese, C. R. (1980). The phylogeny of prokaryotes. Science 209:457-463. ' Gillham, N. W. (1978). Organelle Heredity. Raven Press, New York. pp.602. Gleba, Yu; Komarnitsky, I. K.; Kolesnik. N. N.; Meshkene, 1.; Martyn, G. I. (1985). Transmission genetics of the somatic hybridization process in Nicotiana. II. Plastome heterozyotes. Molec. Gen. Genet. 198:476-481. Gold, B.; Carrillo, N.; Tewari, K. K.; 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. Gordon, K. H. J.; Grouse, E. J.; Bohnert, H. J.; Herrmann, R. G. (1982). Physical mapping of differences in chloroplast DNA of the five wild-type plastomes in Oenothera subsection Fuoenotbera. Theor. Appl. Genet. 61:373-384. 83 Carmen, D. S.; Levine, R. P. (1965). Cytochrome f and plastocynin: their sequence in the photosynthetic electron transport chain of Chlamydomonas reinhardi. Proc. Natl. Acad. Sci. USA 54:1665-1669. Gray, M. W.; Doolittle, W. F. (1982). Has the endosymbiont hypothesis been proven?. Microbiol. Rev. 46:1-42. Grigorieva, G.; Shestakov, S. (1982). Transformation in the cyanobac- terium Synechocystis sp. 6803. FEMS Microbiol. Letters 13:367-370. Hagemann, R. (1976). Plastid distribution and plastid competition in higher plants and the induction of plastome mutations by nitroso-urea- compounds. In: Genetics and Biogenesis of Chloroplasts and Mitochondria (Bacher, Th.; Neupert, W; sebald, W.; and Werner, 8.; eds). Elsevier/ North Holland Press,,Amstsrdam. pp.331-338. Hagemann, R. (1979). Genetics and molecular biology of plastids of higher plants. In: Stadler Symposium Vol 11 (Redei, G.; ed.). Missouri Agricultural Experiment Station. pp.91-115. Hallier, U. W. (1980). Zusammenhange zwischen der genetischen Informa- tion des Plastoms und der Biochemie der Photosynthese. Habilitations- schrift, University of Dusseldorf. Hashimoto, E.; Murakami, S. (1983). Effects of cycloheximide and chloramphenicol on chloroplast replication in synchronously dividing cultured cells of.Eug1en8 graciIis; New Phytol. 94:521-529. Heinhorst, 8.; Cannon, G.; Weissbach A. (1985). Plastid and nuclear DNA synthesis are not coupled in suspension cells of Nicotiaaa tahacum. Plant Mol. Biol. 4:3-12. Hu, NeT.; Thiel, T.; Giddings Jr., T. B.; Wolk, C. P. (1981). New Anabaena and Nastoc cyanophages from sewage settling ponds. Viral. 114:236—246. Kemper, I. (1958). Untersuchungen an panaschierten Genotherea-IBastar- den mit erblich verschiedenen Plastidensorten. Dissertaion Universitfit Koln. Kirk, J. T. O.; Tilnebeassett, R. A. E. (1978). The Plastids. Elsevier/ North Holland. pp.960. Kolodner, R., Tewari, K. K. (1975). Presence of displacement loops in the covalently closed circular chloroplast deoxyribonucleic acid from higher plants. J. Bio. Chem. 250:8840-8847. Kuhlemeier, C. J.; Borrias, W. E.; van den Handel, C. A. M. J. J.; van Arkel, GA (1981). Vectors for cloning in cyanobacteria: construction and characterization of two recombinant plasmids capable of transformation to Escherichia coli K12 and Aaacystis nidulaas R2. Mol. Gen. Genet. 184:249-254. Kutzelnigg, H.; Meyer, B.; Schotz, F. (19758). Untersuchungen an Plas- tommutanten von Oenothera. II. Uberblick iiber die Ultrastruktur der mutierte Plastiden 19758. Biol. Zbl. 94:513-526. Kutzelnigg, H.; Meyer, B.; Schotz, F. (1975b). Untersuchungen an Plas- tomutanten van Oenothera. III. Vergleichende ultrastrukturelle Charak- terisierung der Mutanten 1975b. Biol. Zbl. 94:527-238. Kutzelnigg, H.; Stubbe, W. (1974). Investigation on plastome mutants in Oenothera 1. General considerations. Sub-Cell. Biochem. 3:73-89. Maliga, P; Lorz, H.; Lazar, G.; Nagy, F. (1982). Cytoplast-protoplast fusion for interspecific chloroplast transfer in Nicatiana. Mol. Gen. Genet. 185:211-215. Maniatis, T.; Frisch, E. F.; Samrook, J. (1982). Molecular Cloning. A laboratory manual. Cold Spring Harbor. pp.545. Mascia, P.; Robertson, D. S. (1978). Studies of chloroplast development in four maize mutants defective in chlorophyll biosynthesis. Plants 143: 207-211 . Mayfield, S. P.; Taylor, W. C. (1987). Chloroplast photooxidation in- hibits the expression of a set of nuclear genes. Mol. Gen. Genet. 208:309—314. Madgesy, P.; Fejes, E.; Maliga, P. (1985). Interspecific chlorOplast recodaination in a Nicotiaaa somatic hybrid. Proc. Natl. Acad. Sci. USA 82:6960-6964. Medgyesy, P.; Pay, A.; Merton, L. (1986). Transmission of paternal chloroplasts in Nicotiaaa. Molec. Gen. Genet. 204:195-198. Meta, L. J.; Geist, L. J. (1983). Linkage of a known chloroplast gene mutation to the uniparental genome of Ch18mydamonas reiahardj. Genetics 105:559-579. Metzlaff, M.; Pohlheim, F.; Earner, T.; Hagemann, R. (1982). Hybrid variegation in the genus Pelarg'om’um. Curr. Genet. 5: 245-249. Meyer, B.; Stubbe, W. (1974). Das Zahlenverhéltnis von miitterlichen und vfiterlichen Plastiden in den Zygoten von Oenothera erythrosep818 Borbas (syn. 0e. Ismarckiana). Ber. Deutsch Bot. Ges. 87:29-38. Michaelis, P. (1967). The segregation of plastids as an example of plas- mone analysis. The Nucleus 10:111-127. Murashige, T.; Skoog, F. (1962). A revised medium for rapid growth and bioassays with tobacco tissue cultures. Physiol. Plant 15:473-479. Nagata, T.; Takebe, I. (1971). Plating of isolated tobacco mesophyll protoplasts on agar medium. Planta 88:12-20. Nordstrom, Kurt (1985). Control of plasmid replication: theoretical con- siderations and practical solutions. In: Plasmids in Bacteria (Helinski, D. R.; Cohen, S. N.; Clewell, D. B.; Jackson, D. A.; Hollaender, A.; eds.). Plenum Press, New York, pp. 189-214. Ohtani, T.; Uchimiya, H.; Kato, A.; Harada, H.; Sugita, M. (1984). Loca- tion and nucleotide sequence of a tobacco chloroplast DNA segment capable of replication in yeast. Mol. Gen. Genet. 195:1-4. Ohyama, K.; Fukuzawa, H.; Kohchi, T.; Shirai, H.; Sana, T.; Sano, S.; Umesono, K.; Shiki, Y.; Takeuchi, M.; Chang, Zh.; Aota, S-I.; Inokuchi, H.; Ozeki.H. (1986). Chloroplast gene organization deduced from com- plete sequence of liverwort Marchantia pobmorpha chloroplast DNA. Na- ture 322:572-574. Overbeeke, N.; Haring, M. A.; Nijkamp, H. J. J.; Kool, A. J. (1984). Cloning of Petunia hybridh'chloroplast DNA sequences capable of autonomous replication in yeast. Plant Mol. Biol. 3:235-241. Palmer, J. D. (1985). Comparative organization of chloroplast genomes. Ann. Rev. Genet. 19:325-354. Possingham, J. V.; Lawrence, M. E. (1983). Controls to plastid division. Int. Rev. Cytol. 84:1-56. Ravel-Chapuis, P.; Heizman, P.; Nigon, V. (1982). Electron microscopic localization of the replication origin of Eugiena graciJis chloroplast DNA. Nature 300:78-81. Renner, O. (1915). Befruchtung und Embryobilding bei Oenothera Jamarck- i808 und einigen verwandten arten. Flora 107:115-150. Renner, O. (1924). Die Scheckung der Oenotheredzastarde. Biol. Zentralbl. 44:309-336. Rippka, R.; Deruelles, J.; Waterbury, J. B.; Herdman, M.; Stanier, R. Y. (1979). Generic assignments, strain histories and properties of pure cultures of cyanobacteria. J. Gen. Microbiol. 111:1-61. Rochaix, J. D. (1984). Isolation of chloroplast DNA from Ch1amydosaaas reinhardtii. In: Methods in Chloroplast Molecular Biology, Edelman, M.; Hallick, R. B.; Chua, N-H. (eds). Elsevier Biochemical Press, Amsterdam. pp.295-300. Rochaix, J. D.; van Dillewijn, J.; Rahire, M. (1984). Construction and characterization of autonomously replicating plasmids in the green unicellular alga Ch1amydomoaas reinharoii. Cell 36:925-931. Sager, R. (1972). Patterns of inheritance of organelle genomes: mole- cular basis and evolutionary significance. In: Genetics and Biogenesis of Mitochondria and Chloroplasts (Birky, C. W. Jr.; Perlman, P. S.; 86 Byers, T. J.; eds.). Ohio State University Press, Columbus. pp.252-267. Sager, R.; Kitchen, R. (1975). Selective silencing of eukaryotic DNA. Science 189:426-433. Schiller, B.; Herrmann, R. G.; Melchers, G. (1982). Restriction en- donuclease analysis of plastid DNA from tomato, potato, and some of their somatic hybrids. Mol. Gen. Genet. 186:453-459. Schlunegger, B.; Stutz, E (1984). The Eagleaa grachis chloroplast genome: structural features of a DNA region possibly carrying the single origin of DNA replication. Curr. Genet. 8:629-634. Schmitz, U. K.; Kowallik, K—V. (1986). Plastid inheritance in Epilobium; Curr. Genet. 11:1-6. Schotz, F. (1954). Uber Plastidenkonkurrenz bei Genothera. Plants 43:183-240. Schdtz, F. (1968). Uber plastidenkonkurrenz bei Genothera. II. Biol. Zentralbl . 87: 33-61 . Schotz, F. (1974). Untersuchungen fiber die plastidenkonkurrenz bie Oenothera. IV. Der EinfluB des Genome auf die Durchsetzungsfhhigkeit der Plastiden. Biol. Zentralbl 93:41-64. Schotz, F. (1975). Untersuchungen fiber die plastidenkonkurrenz bie Oenothera. V. Biol. Zentralbl 94:17-26. Schotz, F.; Heiser, F. (1969). Uber Plastidenkonkurrenz bei Oenothera III. Zahlenverhfiltnisse in Mischzellen. Pédagogische Hochschule Potsdam 1:65-89. Scott, N. S., Cain, P., Possingham, V. (1982). Plastid DNA levels in al- bino and green leaves of the mutant of.Hbrdeum vuigare. Z. Pflanzenphysiol. 108:187-192. Scowcroft, W. R.; Larkin, P. J. (1981). Chloroplast DNA assorts randomly in intraspecific somatic hybrids of Nicotiaaa dbhneyi. Theor. Appl. Genet. 60:179-184. Sears, B. B. (1980). Elimination of plastids during spermatogenesis and fertilization in the plant kingdom. Plasmid 4:233-255. Sears, B. B. (1983). Genetics and evolution of the chloroplast. Stadler Symposium V01. 15, Univ. Missouri Press, pp.120-129. Shinozaki, K.; Ohme, M.; Tanaka, M.; Wakasugi, T.; Hayashida, N.; Matsubahashi, T.; Zaita, N.; Chunwongse, J.; Obokata, J.; Yamaguchi- Shinozaki, K.; Ohto, C.; Torazawa, K.; Meng, B. Y.; Sugita, M.; Dena, H.; Kamogashira, T.; Yamada, K.; Kusuda, J.; Takaiwa, F.; Kata, A.; ‘Tondoh, N.; Shimada, H.; Sugiura, M. (1986). The complete nucleotide sequence of the tobacco chloroplast genome: its organization and 87 expression. EMBO J. 5:2043-2049. Struhl, K.; Stinchcomb, D. T.; Scherer, S.; Davis, R. W. (1979). High- frequency transformation of yeast: Autonomous replication of hybrid DNA molecules. Proc. Natl. Acad. Sci. USA 76:1035-1039. Stubbe, W. (1957). Dreifarbenpanaschierung bei Oenothera. I. Entmischung von drei in der zygote vereinigten Plastidomen. Ber. Dtsch Bot. Ges. 70:221-226. Stubbe, W. (1959). Genetische analyse des Zusanmenwirken von Genome und Plastom bei Oenothera. Z. Vererbungsl. 90:288-298. Stubbe, W. (1960). Untersuchungen zur genetischen Analyse des Plastoms von Oenothera. 2. Bot. 48:191-218. Stubbe, W.; Herrmann, R. G. (1982). Selection and maintenance of plas- tome mutants and interspecific genome/plastome hybrids from Oenothera. In: Methods of Chloroplast Molecular Biology (Edelman, M.; Hallick, R.; Chua, N-H.; eds.). Elsevier Biomedical Press, Amsterdam. pp.149-165. Tilney-Bassett, R. A. E. (1986). Plant Chimeras. Edward Arnold Ltd., London. Tilney-Bassett, R. A. E.; Abdel-Wahab, O. A. L. (1979). Maternal effects and plastid inheritance. In: British Society for Developmental Biology Bymposium vol.4. Cambridge university Press. pp.29-45. Tilney-Bassett, R. A. E.; Birky, C. W. (1981). The mechanism of the mixed inheritance of chloroplast genes in Pelargonims. Theor. Appl. Genet. 60:43-53. Vallet, J. M.; Rochaix, J. D. (1985). Chloroplast origins of DNA replication are distinct fromm chloroplast ARS sequences in two green algae. Curr. Genet. 9:321-324. Vaughn, K. C. (1981). Plastid fusion as an agent to arrest sorting out. Curr. Genet. 3:243-245. Weddell, J.; Wang, X. M.; Wu, M. (1984). Electron microscopic localiza- tion of the chloroplast DNA replicative origins in Chlamydomonas reinhardij. Nucleic Acid Res. 12:3843-3865. Walmsley, R. M.; Johnston, L. H.; Williamson, D. H.; Oliver, S. G. (1984). Replicon size of yeast ribosemal DNA. Molec. Gen. Genet. 195:260-266. Whatley, J. M. (1982). Ultrastructure of plastid inheritance: green algae to angiosperms. Biol. Rev. 57:527-569. Whatley, J. M.; Whatley, F. R.; (1981). Chloroplast evolution. New Phytol. 87:233-247. Woese, C. R. (1987). Bacterial evolution. Microb. Rev. 51:221-271. 88 Walk, C. P.; Vonshak, A.; Kehoe, P.; Elhai, J. (1984). Construction of shuttle vectorrs capable of conjugative transfer from Escherichia coIi to nitrogen-fixing filamentous cyanobacteria. Proc. Natl. Acad. Sci. USA 81:1561-1565. Wu, M.; Lou, J. K.; Chang, D. Y.; Chang, 0. H.; Nie, Z. O. (1986). Structure and function of a chloroplast DNA replication origin of Chiasydhmaaas reinhardrii. Proc.‘Natl. Acad. Sci. USA 83:6761-6765. Yamada, T., Shimaji, M., Fukuda, Y. (1986). Characterization of a chloroplast DNA sequence from Ch1ore118 811ipsoidea that promotes autonomous replication in yeast. Plant Mol Biol 6:245-252. Zyskind, J. W.; Cleary, J. M.; Brusilow, W. S. A.; Harding, N. E.; Smith, D. W. (1983). Chromosomal replication origin from the marine bac- terium Vibrio harveyi function in Escherichia co1i: oriC consensus sequence. Proc. Natl. Acad. Sci. USA 80:1164-1168. 293 03046 3305 H " H H "I H Ml U" E“! Ill " H III 3