The Unicellular Green Alga Chlamydomonas reinhardtii as an Experimental System to Study Chloroplast RNA Metabolism

Introduction

The Unicellular Green Alga Chlamydomonas reinhardtii as an Experimental System to Study Chloroplast RNA Metabolism J. Nickelsen, U. Kück (Y) Lehrstuhl für Allgemeine und Molekulare Botanik, Ruhr-Universität Bochum, 44780 Bochum, Germany e-mail: ulrich.kueck6ruhr-uni-bochum.de Tel.: c49-234-3226212 Fax: c49-234-3214184

Chloroplasts are typical organelles of photoautotrophic eukaryotic cells which drive a variety of functions, including photosynthesis. For many years the unicellular green alga Chlamydomonas reinhardtii has served as an experimental organism for studying photosynthetic processes. The recent development of molecular tools for this organism together with efficient methods of genetic analysis and the availability of many photosynthesis mutants has now made this alga a powerful model system for the analysis of chloroplast biogenesis. For example, techniques have been developed to transfer recombinant DNA into both the nuclear and the chloroplast genome. This allows both complementation tests and analyses of gene functions in vivo. Moreover, site-specific DNA recombinations in the chloroplast allow targeted gene disruption experiments which enable a “reverse genetics” to be performed. The potential of the algal system for the study of chloroplast biogenesis is illustrated in this review by the description of regulatory systems of gene expression involved in organelle biogenesis. One example concerns the regulation of trans-splicing of chloroplast mRNAs, a process which is controlled by both multiple nuclear- and chloroplast-encoded factors. The second example involves the stabilization of chloroplast mRNAs. The available data lead us predict distinct RNA elements, which interact with trans-acting factors to protect the RNA against nucleolytic attacks.

Naturwissenschaften 87, 97–107 (2000)

Q Springer-Verlag 2000

The unicellular alga Chlamydomonas reinhardtii displays many of the experimental benefits shared by the budding yeast Saccharomyces cerevisiae. Consequently it has recently been referred to as “green yeast” in a number of reviews to emphasize the technical and organic advantages associated with this organism (Goodenough 1992; Gumpel and Purton 1994; Rochaix 1995). An example of this, one which is also seen in other microorganisms, is that this haploid eukaryot can easily be cultured in inexpensive media, on petri dishes, and in Erlenmeyer flasks. Similar to yeast, meiospore formation is controlled by two mating-type loci and can be used for conventional genetic analysis, for example, for tetrad analysis and recombinant genetics to generate strains with multiple genetic markers. Even more interestingly, the mating-type control is not confined to the inheritance of nuclear but also of extranuclear genes. This enormous advantage for genetic analysis is demonstrated below. However, unlike yeast, C. reinhardtii is flagellated and photosynthetically active, and for this reason the alga has been a preferred experimental organism for studies on flagellate and chloroplast biogenesis, including photosynthesis (Tam and Lefevre 1993; Webber et al. 1995). Figure 1 presents the basic body plan of Chlamydomonas. The ellipsoid C. reinhardtii cell has a polar structure with two anterior flagella, several mitochondria, and a single chloroplast that surrounds the nucleus with the prominent nucleolus. In addition, vacuoles and Golgi vesicles are seen in the cytoplasm. In addition to these features, light microscopy shows an eyespot and pyrenoids surrounded by starch bodies within the chloroplast.

97

Fig. 1. Cell structure of Chlamydomonas reinhardtii. The two- and three-dimensional view of the cell shows the central nucleus (N) with the nucleolus (Nu), the two isoform flagellae (F), the cup-shaped chloroplast (C) with the eyespot (E) and the starch containing pyrenoid (P) and the mitochondria (M). In addition, one may distinguish the golgi vesicle (G), starch grains (S), and the vacuoles (V)

Nuclear, Mitochondrial, and Chloroplast Mutations Can Be Distinguished by Conventional Genetic Analysis Vegetative cells of C. reinhardtii are haploid, and therefore the occurring mutant phenotypes can be directly identified after mutagenesis without the need for further genetic manipulation. For instance, photosynthetic mutants can grow heterotrophically on media containing acetate, which is assimilated in the glyoxylate cycle and utilized as a reduced carbon source (Harris 1989). Another notable characteristic of C. reinhardtii is its capability to synthesize chlorophyll in the dark, allowing the correct assembly of photosynthetic complexes even under heterotrophic conditions. This minimizes indirect, light-induced stress effects during the growth of mutant material (Mets and Rochaix 1998). 98

C. reinhardtii cells grow rapidly, with a doubling time of about 5–8 h depending on the strain and culture conditions used (Harris 1989). Otherwise, minimal growth conditions lead to the onset of the sexual cycle, resulting in the recombination of the genetic material. In the laboratory, gametogenesis can be simply induced by nitrogen deprivation. Isogametes are morphologically similar but differ in their mating type (mt c or mt –), which is determined by a complex genetic locus (Ferris and Goodenough 1994). After flagellar pairing of cells with opposite mating types, a series of events leads to the cytoplasmic fusion of gametes with opposite mating types, which is followed by nuclear and chloroplast fusions. The resulting diploid zygote is protected by a newly synthesized impermeable cell wall. Under appropriate environmental conditions meiosis takes place, resulting in the production of four haploid progeny cells. Their separation and independent maintenance provides the basis for genetic tetrad analysis that allows the inheritance of single mutations to be followed. As with other photosynthetic eukaryotes, C. reinhardtii contains three autonomous genetic systems, which are located in the nucleus, the chloroplast, and the mitochondria. The nuclear genome consists of about 100,000 kbp and represents the vast majority of DNA within the cell. The actual number of existing chromosomes must still be confirmed, but genetic analysis reveals at least 17 linkage groups in C. reinhardtii. As depicted in Fig. 2, nuclear genes segregate according to Mendel in a 2 : 2 ratio among tetrad products during genetic crosses. Thus, when photosynthetic mutants are analyzed, a strict 2 : 2 segregation of photosynthetic activity and deficiency indicates not only that one nuclear gene is affected, but also that only one single or two closely coupled mutations are responsible for the mutant phenotype. The chloroplast (cp) genome of C. reinhardtii has many features in common with higher plants. It is present in approximately 80 copies per chloroplast and comprises a 196 kbp circular molecule containing about 80 genes. Most of these genes encode subunits of either photosynthetic complexes or of the chloroplast transcription/translation machinery. During crosses the chloroplast genome is uniparentally inherited from the mt (plus) parent, leading to a 4 : 0 segregation ratio of chloroplast markers (Fig. 2). The general phenomenon of uniparental inheritance of organelles in eukaryotes can be explained by differently sized gametes, in which chloroplasts and mitochondria are selectively excluded from the smaller gamete, for example, from sperm or pollen cells. However, in the isogametes of C.

Fig. 2. Genetic analysis in Chlamydomonas reinhardtii. While nuclear mutations are inherited in a Mendelian fashion, an uniparental inheritance of chloroplast and mitochondrial genes is observed, depending on the mating type of the parental cells

reinhardtii, other molecular mechanisms, which are not completely understood, guarantee that only cpDNA from the plus mating-type cell is transmitted to the progeny. It appears that during gametogenesis only “plus” cpDNA is reversibly protected by an unknown mechanism. Soon after zygote formation, the “minus” cpDNA is presumably degraded by the action of a nuclease (Kuroiwa 1991). The organization of the C. reinhardtii mitochondrial (mt) genome is clearly different from the chloroplast genome. It consists of multiple copies of a 15.8-kbp linear molecule, which exhibits inverted repeats at its ends that are thought to be involved in mtDNA stabilization. Only eight protein genes, three tRNA genes, and two scrambled rRNA genes are encoded in the mitochondrial genome (Michaelis et al. 1990; Wolff and Kück 1993). In contrast to chloroplasts, mitochondria are inherited uniparentally during ge-

netic crosses from the mt (plus) parent resulting in a 0 : 4 segregation ratio of mitochondrial genes (Fig. 2). Three different types of mitochondrial mutants have been characterized. The complete loss of mitochondrial DNA is associated with the lethal minute phenotype, suggesting that, unlike in yeast mitochondria, some essential functions are encoded in the mt genome. Mutants in the cytochrome pathway of respiration are obligate photoautotrophic (Dorthu et al. 1992; Duby and Matagne 1999). While photosynthesis is dispensable under heterotrophic growth conditions, in C. reinhardtii mitochondrial respiration is dispensable under photoautotrophic growth conditions, enabling the maintenance and characterization of mutants in both processes in one organism. Finally, a number of mutations within the mitochondrial cob gene encoding the apocytochrome b polypeptide have been identified, which confer resistance to the respiration inhibitors myxothiazol and mucidin (Bennoun et al. 1992). These resistances may prove to be used as efficient selectable markers required for genetic engineering of the C. reinhardtii mitochondria by DNA mediated transformation (Randolph-Anderson et al. 1993). In conclusion, the classical tools of conventional genetics enable the inheritance of algal genes to be traced and offer an initial insight into the genetic system with which mutations are associated. Furthermore, mutations in different loci may be combined with appropriate genetic crosses, thereby facilitating the analysis of the effect of double or even triple mutations. Finally, it might be useful for particular biochemical studies to start with “genetically purified” material, for example, with strains in which a component, which normally interferes with the analysis or purification of a different one, is no longer present because its gene is mutated (Pierre et al. 1995).

DNA Transformation and Gene Tagging as Tools for Studying Gene Function Molecular genetics of every organism is highly dependent on reproducible DNA transformation techniques, which can be easily performed in “standard” laboratories. Two different methodologies are currently used for C. reinhardtii to transform nuclear and chloroplast DNA. The so-called “glass bead method” is usually applied when nuclear genes are transferred into the host cell, while particle bombardment is used to integrate recombinant DNA into organellar DNA. 99

Chloroplast Transformation The stable transformation of chloroplasts was long hampered by nonefficient methods to deliver foreign DNA across the cell wall and three membranes, i.e., the plasma membrane and the outer and inner chloroplast envelope membranes. However, with the development of a particle bombardment system this problem has been overcome and C. reinhardtii chloroplasts were the first to be transformed. In the initial successful attempts, DNA-coated particles were accelerated and delivered by powder explosion or by expansion of compressed gas in order to hit cells, which were spread on the surface of an agar plate (Boynton et al. 1988). Further improvements in the particle bombardment technique arrived with the development of the particle inflow gun (Finer et al. 1992), which is illustrated in Fig. 3. DNA-coated tungsten particles are accelerated directly in a helium stream, resulting in reproducible and high transformation rates. Helium release is controlled by a timer relay driven solenoid, which works at low helium pressure and allows optimal particle penetration. The particle flow gun causes less damage to the algal cells than microprojectile delivery systems. Once foreign DNA reaches the chloroplast, it is integrated into the chloroplast genome by homologous recombination. Transformants can be selected on the basis of restored chloroplast defects in photosynthetic functions such as complementation of mutants with deletions in either the atpB or tscA loci (Boynton et al. 1988; Goldschmidt-Clermont et al. 1991). Another approach is to integrate heterologous selectable markers – usually the bacterial aadA gene (Goldschmidt-Clermont 1991) – which confer resistance to antibiotics such as spectinomycin and streptomycin. After integration into one of the chloroplast DNA copies (see above) the foreign DNA segregates rapidly, leading to homoplasmic transformants in which all copies or the chloroplast genome are identical. Using a “reverse genetics” approach, the function of genes and regulatory elements controlling their expression can be further investigated using the biolistic transformation technique to introduce a number of targeted mutations into the coding and noncoding regions of chloroplast genes (for review see Rochaix 1997).

Nuclear Transformation A powerful nuclear transformation system arose when homologous marker genes, such as genes for argininosuccinate lyase (arg7) and nitrate reductase 100

Fig. 3. Schematic illustration of a particle inflow gun to introduce recombinant DNA into the chloroplast genome. Helium pressure (8 bar) is released into the reaction chamber via a syringe filter (F) carrying the DNA coated tungsten particles. Helium release is controlled by a timer relay-driven solenoid (S), which opens the valve for 50 ms. The accelerated particles hit the algal cells, which are spread on the surface of a petri dish (D) on the bottom of the reaction chamber

(NIT1) became available from C. reinhardtii (Purton and Rochaix 1994). Recently also heterologous, dominant, and thus selectable nuclear genes such as the bacterial phleomycin resistance gene (ble) have been used to select for transformed algal cells (Stevens et al. 1996). A further simplification of the technology was achieved when agitation of the cells with glass beads and DNA resulted in the generation of recombinant strains. A prerequisite for use of the “glass bead” method is the availability of a cell wall deficient host strain. Alternatively, the cell wall can be digested with autolysin, an enzyme which is synthesized during mating of C. reinhardtii gametes

with the opposite mating type. In the “glass bead” method usually one or several copies of the transforming DNA are integrated ectopically into the nuclear DNA. Recently an even more efficient transformation method based on electroporation of algal cells has been reported (Shimogawara et al. 1998). There are multiple examples of mutants’ rescues leading to the identification of functional genes. The interested reader is referred to a comprehensive review describing this particular subject matter (Rochaix 1995). Gene tagging is another approach to use DNA transformation for the study of gene and protein function. Here genes of interest are disrupted by targeted insertion of the recombinant DNA. In contrast to conventional UV or chemical mutagenesis, the mutated gene is labeled with foreign DNA and thus can be easily isolated when the DNA sequence of the inserted DNA becomes available. More recently, two significant developments have turned insertion mutagenesis into one of the most successful and important techniques for isolating novel genes from photosynthetic organisms. (a) Several systems based on transposons were substantially improved to perform insertional mutagenesis. For example, transposon tagging has been applied to a wide range of prokaryotic and eukaryotic organisms and also for plant gene identifications. Although several transposons have been described for C. reinhardtii (Day and Rochaix 1991; Ferris 1989; Graham et al. 1995), there are currently no efficient transposon tagging systems available. (b) Alternatively, linearized vector DNA can be used for integrative transformation. The resulting recombinant cells contain one or more molecules of the exogenous DNA ectopically integrated in the nuclear DNA. Random integration of the exogenous DNA results in gene disruption, which is associated with mutant phenotypes. Here we introduce a modified version of this procedure which is called restriction enzyme mediated integration (REMI). This procedure was originally developed for yeasts and slime molds (Kuspa and Loomis 1992; Schiestl and Petes 1991) to generate insertions into genomic restriction sites (Fig. 4A). This occurs when the endonuclease together with the linearized vector molecule is introduced into the host cell. Plasmid DNA is integrated in genomic restriction sites at rather high proportions of the resulting transformants. Using REMI we have constructed a set of C. reinhardtii mutants showing photosynthetic defects. These mutants were identified from REMI transformants by growing the transformed colonies on selective media, which contained or were devoid of acetate as a fixed carbon source (Fig. 4B). Unlike high-

Fig. 4A,B. Construction and selection of nuclear photosynthetic transformants of Chlamydomonas reinhardtii. A) Inactivation of nuclear genes using the REMI procedure. Step A The linearized pStart molecule integrates in restricted genomic DNA of the host cells, resulting in gene disruption. Step B Restriction with an enzyme (R), which cuts outside of the “pStart” sequence, and religation leads to the rescue of recombinant plasmid “pResult.” This can be further characterized, for example, by DNA sequencing. B) Selection scheme to identify photosynthetic mutants among transformants

101

er plants, C. reinhardtii has no need for photosynthesis, and therefore mutants with disrupted photosynthesis are able to grow on acetate-containing plates (Bennoun and Delepelaire 1982). Using these selective growth conditions, REMI transformants can be easily identified as wild type or as photosynthesis mutants solely by growing them on both media. In the next step, the mutant strains were characterized by measurement of their fluorescence induction kinetics (Bennoun and Beal 1998). This represents a noninvasive, in vivo method which allows one to distinguish between defects in particular photosynthetic subcomplexes based on characteristic alterations in fluorescence kinetic curves. From this analysis we identified a subset of mutants with a defective photosystem I, which were further analyzed in DNA hybridization experiments in order to detect the tagged gene within the genomic DNA. Further steps include conventional procedures of recombinant DNA technologies, such as plasmid or phage cloning. An alternative procedure is the PCR technique. After hydrolysis of genomic DNA with endonucleases, religation of restriction fragments facilitates an “inverse PCR” reaction. The amplified DNA fragments can be subjected to cloning experiments, allowing further characterization.

by the chloroplast genome. Hence a lively crosstalk between these two genetic systems must be assumed to be a fundamental condition to guarantee a proper cellular integration of the chloroplast compartment, which is of prokaryotic origin, according to the endosymbiotic hypothesis of organelle evolution. Indeed, accumulating evidence based on the characterization of numerous photosynthetic mutants and comprehensive biochemical analyses indicate that chloroplast gene expression is regulated by nuclear factors that are synthesized on 80S ribosomes in the cytoplasm and subsequently transported into the chloroplast. They then interact with their cognate target sites on chloroplast RNAs or proteins to direct almost all steps of organellar gene expression (Fig. 5). Nuclear gene expression is controlled mainly at the transcriptional level, although genetic evidence suggests that differential RNA stabilization can also play a crucial role for the expression of some nuclear genes encoding chloroplast polypeptides (Hahn et al. 1996). However, for chloroplast gene expression in C. reinhardtii posttranscriptional steps appear to act as control points determining different

Molecular Biology Provides the Basis for Understanding Chloroplast Biogenesis Algae of the genus Chlamydomonas are taxonomically grouped into the order of green algae (Chlorophyceae), which are thought to represent the ancestral prototype of today’s higher plants. It is widely accepted that green algae are sufficiently similar to higher plants to serve as model systems, which can be used to explain the biochemical and molecular aspects of photosynthetic processes that drive biomass production in crops. Although the main steps and components of the light and dark reactions of photosynthesis have been established over the past 20 years, only little is known of the way in which the biogenesis of the photosynthetic machinery is controlled. Obviously, there are constraints opposing a high degree of coordination at the level of gene expression, because cells in both C. reinhardtii and higher plants are challenged with a complex situation. Parts of photosynthetically active complexes, such as those mediating light-driven electron flow across the thylakoid membrane and the CO2 assimilating enzyme Rubisco, are encoded in the nuclear genome, while the remaining subunits are encoded 102

Fig. 5. Chloroplast biogenesis. Photosynthetic multisubunit complexes are formed by both nuclear (black arrows) and chloroplast (white arrows) encoded polypeptides. Coordination of gene expression during chloroplast biogenesis is achieved by nuclear encoded regulatory factors that affect RNA-processing, translation, and finally assembly of the complexes (gray arrows)

levels of gene expression (Rochaix 1996). These steps include the stabilization, 5b and 3b end processing, splicing, and translation of chloroplast transcripts and, furthermore, posttranslational modifications of polypeptides and the final assembly of photosynthetic multisubunit complexes (Fig. 5). Photosynthetic algal mutants show defects in almost all of these posttranscriptional events. An intriguing feature of the underlying mutations is that they act in a gene-specific manner affecting the expression of single chloroplast genes only. This stands in contrast to similar mutations in higher plants, which usually exhibit more pleiotropic effects (Barkan et al. 1995; Dinkins et al. 1997; Meurer et al. 1996). The underlying reasons for this difference have not yet been fully explained, but it is likely that pleiotropic mutations affecting all chloroplast functions are lethal in the alga (Goldschmidt-Clermont 1998). This suggests that the expression of at least a few chloroplast genes is required for cell survival. In higher plants chloroplast-encoded functions are dispensable, as suggested by the viability of chloroplast-ribosome deficient mutants such as iojap in maize and albostrians in barley (Börner and Sears 1986). Thus the usual screens which are based on high chlorophyll fluorescent phenotypes of cells deficient in photosynthetic electron flow (see above) would select only those mutants in C. reinhardtii that are affected in the group of gene-specific, nuclearencoded factors (Goldschmidt-Clermont 1998). However, the combined analysis of nuclear photosynthetic and site-directed chloroplast mutants as well as chloroplast transformants carrying reporter genes provides a powerful approach for the study of chloroplast biogenesis in C. reinhardtii at the molecular level. A recent book covers various aspects of chloroplast and mitochondrial biogenesis with a set of comprehensive reviews (Rochaix et al. 1998). We therefore focus in the following section on current knowledge about the regulation of posttranscriptional chloroplast gene expression by providing supporting examples illustrating the processes of splicing and stabilization of plastid transcripts.

Splicing of Chloroplast Introns Some chloroplast genes are interrupted by noncoding intervening sequences called introns. These are divided into two main subclasses, namely group I and group II, based on conserved primary and secondary structural elements (Michel and Dujon 1983). Five group I introns have now been identified in C. reinhardtii chloroplasts. One is located within

the rrnL gene encoding the chloroplast 23S rRNA and the remaining four disrupt the psbA gene encoding the D1 protein of photosystem II. All five introns exhibit autocatalytic self-splicing activity under appropriate nonphysiological in vitro conditions. However, it is thought that yet unidentified transacting factors assist the splicing reaction in vivo (Herrin et al. 1990, 1991). Interestingly, splicing of all four psbA introns is regulated by light and requires photosynthetic electron transport suggesting that psbA gene expression is controlled, at least partially, at the level of RNA maturation (Deshpande et al. 1997). Group II introns are defined by a conserved secondary structure composed of six partly helical domains surrounding a central wheel. Two transesterification reactions lead to the excision of the intron and ligation of the exons, and, similar to group I introns, also some group II introns show self-splicing activity in vitro. No continuous group II introns have yet been detected in C. reinhardtii, but the psaA gene encoding the P700 apoprotein of photosystem I has been found to be separated into three independently transcribed exons, each flanked by sequences that share some of the conserved structural features of group II introns (Kück et al. 1987; Fig. 6). This unusual organization suggests that psaA RNA maturation is achieved via two trans-splicing reactions. In addition, another chloroplast locus, called tscA, is involved in the first trans-splicing step fusing psaA exon 1 and exon 2. The tscA gene encodes a small RNA which has no protein-coding capacity but appears to complement the first psaA intron structure by providing the precursor transcripts of exon 1 and exon 2 with the helical domains I–IV (GoldschmidtClermont et al. 1991). Thus, the first psaA intron is composed of at least three different RNA molecules. This finding has led to speculation as to whether the split psaA intron 1 structure reflects some early steps in the evolution of nuclear introns. Their splicing is also dependent on trans-acting RNAs, the so-called snRNAs, which share structural similarities with organellar group II introns (Sharp 1991). Thorough genetic analyses identified a large number of additional nucleus-encoded functions that are required for the complex psaA maturation steps. At least 14 different nuclear genes were identified, which can be grouped into three classes. In class C mutants such as M18 and TR72 (Fig. 6), exclusively trans-splicing of exons 1 and 2 is blocked. Conversely, in class A mutants exons 1 and 2 are spliced, but not exons 2 and 3. Finally, class B mutants such as L118 and HN31 (Fig. 6) fail to splice both introns (Goldschmidt-Clermont et al. 1990). Not all of the 103

Fig. 6. Trans-splicing of the tripartide psaA gene, encoding the P700 chlorophyl a/b binding protein of photosystem I. Three precursor mRNA participate in the formation of the continuous PsaA encoding mRNA. The trans-splicing process is mediated by a chloroplast (cpDNA) encoded tscA RNA and several nuclear factors (circled) which are further described in the text

affected factors are necessarily involved in the splicing reaction directly but might act more indirectly, as is indicated by current data on the molecular characterization of some mutants. In the class C mutant TR72 no mature tscA RNA is detected. Instead, high molecular weight precursor RNAs also containing the cotranscribed chloroplast chlN gene accumulate, indicating that the 3b end processing of tscA RNA is defective (Hahn et al. 1998; Fig. 6). Thus the splicing deficiency in TR72 appears to be an indirect effect of the missing maturation of the splicing cofactor tscA. However, it cannot be ruled out that the TR72 gene product is also involved in the downstream splicing reaction. It is anticipated that the recent cloning of the TR72 locus will help to clarify this matter (J. Nickelsen, D. Hahn, V. Holländer, C. Ollbrich, and U. Kück, unpublished results). The situation is even more complicated for the class B mutant HN31. The HN31 gene product seems to be required for both correct tscA 3b end processing and the splicing of psaA exons 2 and 3 (Fig. 6). Backcrosses of HN31 with the wild type revealed that only a single locus is involved in these 104

two different RNA processing events suggesting the involvement of a bifunctional factor (Hahn et al. 1998). The ongoing cloning of the genes affected in various mutants will facilitate the analysis of the molecular mechanics promoting trans-splicing of the psaA message and will extend the knowledge of the way in which splicing mechanisms have developed during evolution. As noted above, no continuous group II introns have been found in C. reinhardtii chloroplast genes. It was therefore interesting to test whether such group II introns from other organelles can be spliced after their biolistic introduction into the algal chloroplast genome. While the atpF group II intron from spinach chloroplasts was not spliced (Deshpande et al. 1995), the mitochondrial rI1 intron from Scenedesmus obliquus was excised accurately (Herdenberger et al. 1994). These differences remain to be investigated, but it may be that subtle structural differences – the two introns form parts of different intron subgroups – cause the difference in ability to be spliced. The efficient rI1 splicing activity enabled analysis of the effect of site-directed mutations within the intron sequence in vivo, leading to a detailed map of sequence elements that are critical for or affect the splicing reaction (Holländer and Kück 1998, 1999a). Interestingly, in the eubacterium Escherichia coli the same intron derivatives are spliced in a similar manner as in chloroplast splicing; thus it is likely that similar ubiquitous factors are involved in the two organisms (Holländer and Kück 1999b).

RNA Stabilization The steady-state level of a given RNA is determined by two processes, its synthesis and its degradation. In chloroplasts, in contrast to the nuclear compartment, the main control appears to be mediated by differential RNA stabilization events, although significant changes in the transcription rates of some chloroplast genes during the cell cycle of C. reinhardtii have been reported (Salvador et al. 1993a). The current models of chloroplast RNA stabilization predict distinct RNA elements, which interact with trans-acting factors to protect the RNA against nucleolytic attack (Nickelsen 1998). By analyzing several nuclear mutants which exhibit defects in the accumulation of single chloroplast transcripts and by using the above chloroplast transformation techniques, involved factors and target sites on chloroplast transcripts can now be identified. The introduction of site-directed chloroplast mutations has revealed that both the 5b and the 3b un-

translated regions (UTRs) of chloroplast messages can contain elements required for the stability of the corresponding transcripts. For instance, the deletion of so-called stem-loop structures located within the 3b UTRs of the atpB and psaB mRNAs leads to a drastic reduction in the accumulation of these transcripts (Lee et al. 1996; Stern et al. 1991). On the other hand, 5b regions have been shown to be crucial for the stabilization of the rbcL mRNA in both C. reinhardtii and tobacco (Salvador et al. 1993b; Shiina et al. 1998). Furthermore, at least two nuclear C. reinhardtii mutants, namely nac2-26 and F16, have been reported to destabilize the psbD and petD mRNA, respectively, via the 5b UTR (Drager et al. 1998; Nickelsen et al. 1994). Apparently, a factor is affected in both mutants that protects the corresponding transcripts against 5bP3b directed exonucleolytic degradation. The expression of the psbD gene encoding the reaction center D2 protein of photosystem II has been investigated by detailed analysis of site-directed mutants of the psbD 5b UTR (Nickelsen et al. 1999). Two cis-elements required for psbD RNA stabilization have been mapped, one within the first 12 nucleotides of the psbD leader and a second one around position P30 relative to the AUG start codon (Fig. 7, I and II, respectively). In addition, three distinct elements involved in the translation of this message have been found. One comprises a U-rich tract immediately downstream of the second RNA stability element, a sequence resembling a prokaryotic Shine-Dalgarno motif (GGAG) and, as expected, the AUG initiation codon (Fig. 7, III, IV and AUG, respectively). The situation is further complicated by the fact that the psbD 5b UTR, as with several other C. reinhardtii chloroplast mRNAs, undergoes a 5b processing reaction. This

maturation appears to be tightly coupled to translation initiation, as suggested by the distribution of the precursor and the mature psbD RNA form on polysome profiles (Nickelsen et al. 1999). Moreover, 5b processing appears to depend on the Nac2 factor affected in the psbD RNA stability mutant nac2-26. This is based on experiments in which the psbD mRNA was artificially stabilized in the nuclear nac2-26 background by inserting a stretch of 18 consecutive G residues into the 5b UTR. Such G tracts have previously been shown to stabilize atpB and petD transcripts in C. reinhardtii chloroplasts (Drager et al. 1996, 1998). Despite the stable accumulation of psbD transcripts in this G tract containing, nonphotosynthetic strain, no 5b RNA processing or D2 synthesis was observed, suggesting a close connection between processes of RNA stabilization, 5b maturation, and translational initiation. Further evidence of such a connection has been obtained through recent data indicating that a 40-kDa protein interacting specifically with the translational U-rich element binds this region in a Nac2-dependent manner (Ossenbühl and Nickelsen, unpublished results). The recent identification of the Nac2 locus after complementation of the mutant and a subsequent plasmid rescue approach classifies the encoded polypeptide into a group of so-called tetratrico-peptide proteins. These are thought to mediate protein-protein interactions and thereby multisubunit complex formations (Nickelsen and Rochaix, unpublished results). In conclusion, an overall picture of gene specific stabilization effects is emerging, in which defined cis-acting elements on chloroplast RNAs interact with their cognate trans-acting factors to determine particular RNA steady-state levels. This process can be closely linked to downstream steps of chloroplast gene expression, such as 5b RNA maturation and translation initiation and, additionally, might be mediated by large multisubunit complexes that successively process the mRNAs up to the final synthesis and assembly of the encoded proteins.

Conclusion

Fig. 7. Control of psbD gene expression. The nuclear encoded Nac2 factor is required for 5b processing of the psbDmRNA and its stabilization. Only the processed mRNA can be translated in the chloroplasts. Regulatory RNA elements of the psbD 5b UTR, which are required for other RNA stabilization (open boxes) or translation (dotted boxes), are indicated. Roman numerals See text

The availability of efficient nuclear and chloroplast transformation systems for C. reinhardtii, along with other techniques of molecular genetics which are very similar to those available in yeast, provides plant researchers with a highly valuable system for studying chloroplast biogenesis. The tools available and the study of photosynthesis mutants from this “green yeast” have provided a complex picture of 105

organelle biogenesis in a plant model system. Several genes controlling chloroplast biogenesis have been characterized, and it has therefore become possible to identify regulatory components of chloroplast biogenesis involved in RNA maturation and stabilization in detail. We thank Ms. Inés Langner for typing the manuscript, Mr. H.J. Rathke for the excellent artwork, Dr. M. Nowrousian for an early version of Fig. 4A, and Prof. P. Bennoun (Paris) for his long-standing collaboration in the mutant analysis. The experimental work of the authors is funded by individual grants from the Deutsche Forschungsgemeinschaft (Bonn–Bad Godesberg, Germany).

Barkan A, Voelker R, Mendel-Hartvig J, Johnson D, Walker M (1995) Genetic analysis of chloroplast biogenesis in higher plants. Physiol Plant 93 : 163–170 Bennoun P, Beal D (1998) New digital imaging instrument for measuring fluorescence and delayed luminescence. In: Rochaix J-D, Goldschmidt-Clermont M, Merchant S (eds) The molecular biology of chloroplasts and mitochondria in Chlamydomonas. Kluwer, Dordrecht, pp 451–458 Bennoun P, Delepelaire P (1982) Isolation of photosynthesis mutants in Chlamydomonas. In: Edelman M, Hallick RB, Chua NH (eds) Methods in chloroplast molecular biology. Elsevier, Amsterdam, pp 25–38 Bennoun P, Delosme M, Kück U (1992) New tools for mitochondrial genetics of Chlamydomonas reinhardtii: manganese mutagenesis and cytoduction. Mol Gen Genet 234 : 147–154 Börner T, Sears BB (1986) Plastome mutants. Plant Mol Biol Rep 4 : 69–92 Boynton JE, Gillham NW, Harris EH, Hosler JP, Johnston AR, Jones BL, Randolph-Anderson D, Robertson TM, Klein KB, Shark B, Sanford J (1988) Chloroplast transformation in Chlamydomonas with high velocity microprojectiles. Science 240 : 1534–1538 Day A, Rochaix JD (1991) Conservation in structure of TOC1 transposons from Chlamydomonas reinhardtii. Gene 104 : 235–239 Deshpande NN, Hollingsworth M, Herrin DL (1995) The atpF groupII intron-containing gene from spinach chloroplasts is not spliced in transgenic Chlamydomonas chloroplasts. Curr Genet 28 : 122–127 Deshpande NN, Bao Y, Herrin DL (1997) Evidence for light/redox regulated splicing of the psbA introns in Chlamydomonas chloroplasts. RNA 3 : 37–48 Dinkins RD, Bandaranayake H, Baeza L, Griffiths AJ, Green BR (1997) hcf5, a nuclear photosynthetic electron transport mutant of Arabidopsis thaliana with a pleiotropic effect on chloroplast gene expression. Plant Physiol 113 : 1023–1031 Dorthu MP, Remy S, Michel-Wolwertz MR, Colleaux L, Breyer D, Beckers MC, Engelbert S, Duyckaerts C, Sluse FE, Matagne RF (1992) Biochemical, genetic and molecular characterization of new respiratory-deficient mutants in Chlamydomonas reinhardtii. Plant Mol Biol 18 : 759–772 Drager RG, Zeidler M, Simpson CL, Stern DB (1996) A chloroplast transcript lacking the 3b inverted repeat is degraded by 3bP5b exoribonuclease activity. RNA 2 : 652–663 Drager RG, Girard-Bascou J, Choquet Y, Kindle KL, Stern DB (1998) In vivo evidence for 5bP3b exoribonuclease degradation of an unstable chloroplast mRNA. Plant J 13 : 85–96 Duby F, Matagne RF (1999) Alteration of dark respiration and reduction of phototrophic growth in a mitochondrial DNA deletion mutant of Chlamydomonas lacking cob, nd4, and the 3b end of nd5. Plant Cell 11 : 115–126 Ferris PJ (1989) Characterization of a Chlamydomonas transposon, Gulliver, resembling those in higher plants. Genetics 122 : 363–377 106

Ferris PJ, Goodenough UW (1994) The mating-type locus of Chlamydomonas reinhardtii contains highly rearranged DNA sequences. Cell 76 : 1135–1145 Finer JJ, Vain P, Jones MW, McMullen MD (1992) Development of the particle inflow gun for DNA delivery to plant cells. Plant Cell Rep 11 : 323–328 Goldschmidt-Clermont M (1998) Coordination of nuclear and chloroplast gene expression in plant cells. Int Rev Cytol 177 : 115–180 Goldschmidt-Clermont M (1991) Transgenic expression of aminoglycoside adenine transferase in the chloroplast: a selectable marker for site-directed transformation of Chlamydomonas. Nucleic Acids Res 19 : 4083–4089 Goldschmidt-Clermont M, Girard-Bascou J, Choquet Y, Rochaix J-D (1990) Trans-splicing mutants of Chlamydomonas reinhardtii. Mol Gen Genet 223 : 417–425 Goldschmidt-Clermont M, Choquet Y, Girard-Bascou J, Michel F, Schirmer-Rahire M, Rochaix J-D (1991) A small chloroplast RNA may be required for trans-splicing in Chlamydomonas reinhardtii. Cell 65 : 135–143 Goodenough UW (1992) Green Yeast. Cell 70 : 533–538 Graham JE, Spanier JG, Jarvik JW (1995) Isolation and characterization of Pioneer1, a novel Chlamydomonas transposable element. Curr Genet 28 : 429–443 Gumpel NJ, Purton S (1994) Playing tag with Chlamydomonas. Trends Cell Biol 4 : 299–301 Hahn D, Bennoun P, Kück U (1996) Altered expression of nuclear genes encoding chloroplast polypeptides in non-photosynthetic mutants of Chlamydomonas reinhardtii: evidence for post-transcriptional regulation. Mol Gen Genet 252 : 362–370 Hahn D, Nickelsen J, Hackert A, Kück U (1998) A single nuclear locus is involved in both chloroplast RNA trans-splicing and 3b end processing. Plant J 15 : 575–581 Harris EH (1989) The Chlamydomonas sourcebook. Academic, San Diego Herdenberger F, Holländer V, Kück U (1994) Correct in vivo RNA splicing of a mitochondrial intron in algal chloroplasts. Nucleic Acids Res 22 : 2869–2875 Herrin DL, Chen Y-F, Schmidt GW (1990) RNA splicing in Chlamydomonas chloroplasts: self-splicing of 23S preRNA. J Biol Chem 265 : 21134–21140 Herrin DL, Bao Y, Thompson AJ, Chen Y-F (1991) Self-splicing of the Chlamydomonas chloroplast psbA introns. Plant Cell 3 : 1095–1107 Holländer V, Kück U (1998) Splicing of the mitochondrial group II intron rI1 : conserved intron-exon interactions diminish splicing efficiency. Curr Genet 33 : 117–123 Holländer V, Kück U (1999a) Group II intron splicing in chloroplasts: identification of mutations determining intron stability and fate of exon RNA. Nucleic Acids Res 27 : 2345–2353 Holländer V, Kück U (1999b) Group II intron splicing in Escherichia coli: phenotypes of cis-acting mutations resemble splicing defects observed in organelle RNA processing. Nucleic Acids Res 27 : 2339–2344 Kück U, Choquet Y, Schneider M, Dron M, Bennoun P (1987) Structural and transcription analysis of two homologous genes for the P700 chlorophyll a-apoproteins in Chlamydomonas reinhardtii: evidence for in vivo trans-splicing. EMBO J 6 : 2185–2195 Kuroiwa T (1991) The replication, differentiation, and inheritance of plastids with emphasis on the concept of organelle nuclei. Int Rev Cytol 128 : 1–62 Kuspa A, Loomis WF (1992) Tagging developmental genes in Dictyostelium by restriction enzyme-mediated integration of plasmid DNA. Proc Natl Acad Sci USA 89 : 8803–8807 Lee H, Bingham SE, Webber AN (1996) Function of 3b non-coding sequences and stop codon usage in expression of the chloroplast psaB gene in Chlamydomonas reinhardtii. Plant Mol Biol 31 : 337–354

Mets LJ, Rochaix J-D (1998) Perspectives. In: Rochaix J-D, Goldschmidt-Clermont M, Merchant S (eds) The molecular biology of chloroplasts and mitochondria in Chlamydomonas. Kluwer, Dordrecht, pp 685–703 Meurer J, Berger A, Westhoff P (1996) A nuclear mutant of Arabidopsis with impaired stability on distinct transcripts of the plastid psbB, psbD/C, ndhH and ndhC operons. Plant Cell 8 : 1193–1207 Michaelis G, Vahrenholz C, Pratje E (1990) Mitochondrial DNA of Chlamydomonas reinhardtii: the gene for apocytochrome b and the complete functional map of the 15.8 kb DNA. Mol Gen Genet 223 : 211–216 Michel F, Dujon B (1983) Conservation of secondary structures in two intron families including mitochondrial-, chloroplast- and nuclear-encoded members. EMBO J 2 : 33–38 Nickelsen J (1998) Chloroplast RNA stability. In: Rochaix J-D, Goldschmidt-Clermont M, Merchant S (eds) The molecular biology of chloroplasts and mitochondria in Chlamydomonas. Kluwer, Dordrecht, pp 151–163 Nickelsen J, van Dillewijn J, Rahire M, Rochaix J-D (1994) Determinants for stability of the chloroplast psbD RNA are located within its short leader region in Chlamydomonas reinhardtii. EMBO J 13 : 3182–3191 Nickelsen J, Fleischmann M, Boudreau E, Rahire M, Rochaix J-D (1999) Identification of cis-acting RNA leader elements required for chloroplast psbD gene expression in Chlamydomonas. Plant Cell 11 : 957–970 Pierre Y, Breyton C, Kramer D, Popot J-L (1995) Purification and characterization of the cytochrome b6f complex of Chlamydomonas reinhardtii. J Biol Chem 49 : 29342–29349 Purton S, Rochaix JD (1994) Complementation of a Chlamydomonas reinhardtii mutant using a genomic cosmid library. Plant Mol Biol 24 : 533–537 Randolph-Anderson BL, Boynton JE, Gillham NW, Harris EH, Johnson AM, Dorthu MP, Matagne RF (1993) Further characterization of the respiratory deficient dum-1 mutation of Chlamydomonas reinhardtii and its use as a recipient for mitochondrial transformation. Mol Gen Genet 236 : 235–244 Rochaix JD (1995) Chlamydomonas reinhardtii as the photosynthetic yeast. Annu Rev Genet 29 : 209–230

Rochaix JD (1996) Post-transcriptional regulation of chloroplast gene expression in Chlamydomonas reinhardtii. Plant Mol Biol 32 : 327–341 Rochaix JD (1997) Chloroplast reverse genetics: new insights into the function of plastid genes. Trend Plant Sci 2 : 419–425 Salvador ML, Klein U, Bogorad L (1993a) Light regulated and endogenous fluctuations of chloroplast transcript levels in Chlamydomonas. Regulation by transcription and RNA degradation. Plant J 3 : 213–219 Salvador ML, Klein U, Bogorad L (1993b) 5b sequences are important positive and negative determinants of the longevity of Chlamydomonas chloroplast gene transcripts. Proc Natl Acad Sci USA 90 : 1556–1560 Schiestl RH, Petes TD (1991) Integration of DNA fragments by illegitimate recombination in Saccharomyces cerevisiae. Proc Natl Acad Sci USA 88 : 7585–7589 Sharp PA (1991) Five easy pieces. Science 254 : 663 Shiina T, Allison L, Maliga P (1998) RbcL transcript levels in tobacco plastids are independent of light: reduced dark transcription rate is compensated by increased mRNA stability. Plant Cell 10 : 1713–1722 Shimogawara K, Fujiwara S, Grossman A, Usuda H (1998) High-efficiency transformation of Chlamydomonas reinhardtii by electroporation. Genetics 148 : 1821–1828 Stern DB, Radwanski ER, Kindle KL (1991) A 3b stem/loop structure of the Chlamydomonas chloroplast atpB gene regulates mRNA accumulation in vivo. Plant Cell 3 : 285–297 Stevens Dr, Rochaix JD, Purton S (1996) The bacterial phleomycin resistance gene ble as a dominant selectable marker in Chlamydomonas. Mol Gen Genet 251 : 23–30 Tam L-W, Lefebvre PA (1993) Cloning of flagellar genes in Chlamydomonas rheinhardtii by DNA insertional mutagenesis. Genetics 135 : 375–384 Webber AN, Scott EB, Hyeonmoo L (1995) Genetic engineering of thylakoid protein complexes by chloroplast transformation in Chlamydomonas rheinhardtii. Photosynthesis Res 44 : 191–205 Wolff G, Kück U (1993) Organization and coding capacity of mitochondrial genomes of algae. In: Brennicke A, Kück U (eds) Plant mitochondria. VCH, Weinheim, pp 103–113

107