Metabolic Engineering via the Chloroplast Genome PDF

CHAPTER 3 PLASTID PATHWAYS Metabolic engineering via the chloroplast genome TRACEY RUHLMAN AND HENRY DANIELL Department of Molecular Biology & Microbiology, University of Central Florida, Biomolecular Science, Building #20, Room 336, Orlando, FL 32816-2364, USA Abstract: Plant metabolic engi...

CHAPTER 3 PLASTID PATHWAYS Metabolic engineering via the chloroplast genome TRACEY RUHLMAN AND HENRY DANIELL Department of Molecular Biology & Microbiology, University of Central Florida, Biomolecular Science, Building #20, Room 336, Orlando, FL 32816-2364, USA Abstract: Plant metabolic engineering has the potential to provide for the needs of an expanding population. Environmentally benign biosyntheisis of novel materials and pharmaceu- tical proteins along with the opportunity to improve the productivity and nutritive value of crop plants has focused considerable effort towards the genetic manipulation of crop species. The most important output traits that could be conferred through biotech- nology often require the coordinated expression of several foreign genes. Conservative estimates predict some 3000 proteins are posttranslationally imported into plant plastids. Among them are the enzymes of various metabolic pathways such as those involved in the biosynthesis of the tocopherols (vitamin E) and carotenoids (vitamin A), branched chain and aromatic amino acids, and fatty acids. The ability of the chloroplast to integrate and express foreign sequences as operons makes this site an attractive alter- native for genetic manipulations. Multigene engineering, high levels of recombinant protein accumulation and the security of transgene containment due to maternal inher- itance of plastid genomes in most crop species are some of the features that contribute to the potential of the chloroplast system. Here we offer an overview of the funda- mental characteristics of plastid protein expression and consider some possible candidate genes for the improvement of crop species through metabolic engineering of pathways compartmentalized within plastids Keywords: chloroplast, oral delivery, genetic transformation, operon, UTR, protein expression, nutritional enhancement To whom correspondence should be addressed. Henry Daniell, Pegasus Professor & Trustee Chair, University of Central Florida, 4000 Central Florida Blvd, Department Molecular Biology & Microbi- ology Biomolecular Science, Bldg # 20, Room 336, Orlando FL 32816-2364, Phone: 407-823-0952, Fax: 407-823-0956, E-mail: [email protected] 79 R. Verpoorte et al. (eds.), Applications of Plant Metabolic Engineering, 79–108. © 2007 Springer. 80 RUHLMAN AND DANIELL 1. INTRODUCTION The current global population of 6.4 billion is expected to reach 10 billion by the year 2050. The rate of agricultural yield at present is not sufficient to meet this demand and already malnutrition and starvation are taking a toll world wide. In the past, productivity of primary producers, namely higher plants, has been accomplished through selective breeding programs but the successes in this area have reached a plateau. Bringing new acreage under cultivation is not a viable option as much of the unutilized lands in developing nations is of marginal quality and serious environmental consequences prohibit agricultural development on remaining fertile preserves. The architects of the ‘green revolution’ envisioned that increased global carrying capacity would result from the development of new crop cultivars, the use of irrigation systems, and the application of chemical fertilizers and pesticides. Food production increased over 1000% from 1960 to 1990 but not without consequences in terms of production cost, dependence on chemical inputs, top soil erosion and salinization due to heavy fertilizer use and the development of pesticide-resistant species. Nobel Peace Prize winner Norman Borlaug, considered the father of the green revolution, suggests that in biotechnology lies the potential to ameliorate environ- mental concerns, while meeting the rising demand for agricultural production (Borlaug, 2005). Indeed, through biotechnology many improvements have been made. Crop species have been genetically engineered to resist viral pathogens (Fitch et al., 1992) and insect pests (Perlak et al., 1990), tolerate drought (Shou et al., 2004) and herbicide treatment (Padgette et al., 1995; Chin et al., 2003), and to enhance nutritional value (Goto et al., 1999; Ye et al., 2000; Datta et al., 2003; Baisakh et al., 2006) through the incorporation of novel DNA into the nuclear genome. Specific technical challenges related to the random integration of foreign DNA sequences such as transgene silencing (Fagard and Vaucheret, 2000) are further compounded by negative public sentiment partially fueled by the fear of transgene escape via pollen or seeds. Without the continued acceptance of regulatory agencies and the general public, recent advances made through biotechnology will be of limited use, and unable to facilitate much needed improvements in the field. The responsibility of finding new platforms for increased agricultural productivity without concomitant threat to the environment falls to science. 1.1. An Alternative Biotechnology Concept Crop plants possess two genomes in addition to that of the nucleus, the organellar genomes of mitochondria and chloroplasts. Genetic engineering of higher plant chloroplasts may offer the potential to mitigate certain limitations of agricultural productivity. Technological advances, most notably the invention of the particle accelerator (Boynton et al., 1988), and the ability to express foreign genes in plastids (Daniell et al., 1987; 1990) , have provided the opportunity to explore the chloroplast genome as a new platform to address current and future demands for improved food production. The concept of chloroplast genetic engineering has been demonstrated  PLASTID PATHWAYS 81 to confer desirable plant traits including insect resistance (McBride et al., 1995; De Cosa et al., 2001), herbicide resistance (Daniell et al., 1998; Iamtham and Day 2000), salt tolerance (Kumar et al., 2004a), drought tolerance (Lee et al., 2003), disease resistance (De Gray et al., 2001), phytoremediation (Ruiz et al., 2003) and reversible male sterility (Ruiz and Daniell, 2005). 1.1.1. Pharming is the future Beyond the ability to increase the abundance and nutritional value of food sources lies the potential to express proteins of industrial interest in plant plastids. Accumu- lation of such diverse compounds as liquid crystal polymers (Viitanen et al., 2004), vaccine antigens (Daniell et al., 2005b; Koya et al., 2005; Chebolu and Daniell, 2007) and other clinically relevant proteins (Staub et al., 2000; Guda et al., 2000; De Gray et al., 2001; Fernandez-San Millan et al., 2003; Leelavathi and Reddy 2003, Arlen et al., 2007; Ruhlman et al., 2007) has been achieved without concomitant loss of host plant viability. Concentrations at or beyond that which would be required for feasibility in subsequent refinement procedures have been obtained. Moreover the possibility to deliver pharmaceutical proteins orally has far reaching implica- tions. Elimination of costs related to purification, cold shipping and storage and the need for sterile injection by medical professionals make the advancement of this technology very attractive. Local production of the source crop would be an additional advantage to developing nations adding to the allure of these systems as pharmaceutical platforms. The very exciting recent accomplishment of transmu- cosal delivery of a plant produced cholera toxin subunit B (CTB) green fluorescent protein (GFP) fusion via the ganglioside M1 (GM1) receptor on the cells of the intestinal epithelia (Limaye et al., 2006) demonstrates the potential of a plant based oral delivery system. Inclusion of the recognition sequence for the ubiquitous protease furin between CTB and GFP allowed for intracellular cleavage of the fusion product and subsequent transport of GFP, but not CTB, via the mucosal vasculature to the liver and spleen of mice fed with pulverized transgenic leaf tissue (Limaye et al., 2006). Ongoing work in our lab employs the CTB/GM1 system to orally deliver plastid produced human proinsulin, as a fusion with CTB, to non-obese diabetic (NOD) mice. Our results indicate that presentation of the chloroplast derived antigen in the gut associated lymphoid tissue can prevent pancreatic insulitis in the NOD model (Devine 2005, Ruhlman et al., 2007) supporting the application of this technology to the treatment of clinical conditions. Tobacco plants expressing the CTB-proinsulin fusion (CTB-Pins) accumulated foreign protein up to 16% of TSP in leaf tissue. Elevated expression of immunosuppressive cytokines such as inter- leukin 4 and 10 was observed following oral gavage of NOD mice with low doses of tobacco leaf material. Concomitantly significant preservation of the pancreatic islets of Langerhans was seen in NOD mice fed CTB-Pins expressing tobacco compared to controls. Furthermore we have demonstrated that the CTB-Pins fusion protein can be produced in lettuce chloroplasts via stable transformation of the lettuce plastome. CTB-Pins accumulated up to 2.5% of TSP in lettuce leaves when expression was driven by the tobacco plastid ribosomal operon promoter and 82 RUHLMAN AND DANIELL the 5’ translational control region of bacteriophage T7 gene 10 (T7g10; Ruhlman et al., 2007). The accumulation of a pharmaceutically important protein such as CTB-Pins in non-toxic lettuce leaves brings us closer to the accomplishment of orally delivered therapies. 1.1.2. The protein producing potential of plastids The genetic potential of the chloroplast lies in its negatively supercoiled, double stranded, circular DNA molecule referred to as the plastome. Within the angiosperms the plastome carries approximately 120 to 130 genes and ranges in size from 120 to 180 kilobases (kb) (Sugiura, 1992). Plastome molecules are clustered in nucleiods which are associated with plastid membranes and readily observed by fluorescent microscopy following DAPI staining (Mache and Lerbs-Mache, 2001). Of the estimated 3000 or so proteins found in the higher plant chloroplast (Colas des Francs-Small et al., 2004; Richly and Leister, 2004), only a small fraction are encoded by the plastome. The bulk of the chloroplast proteome is nuclear encoded, translated on cytosolic ribosomes and subsequently translocated across the chloroplast envelopes (Zerges, 2000). The plastome exists in a highly polyploid state with up to 100 identical copies present in each plastid of a mature leaf cell (Maier et al., 2004). In a mature leaf, mesophyll cells carry up to 100 chloroplasts with the result that this genome alone can comprise up to 20% of the total cellular DNA content (Bendich, 1987). The plastome persists in all plastid differentiation types: the proplastids of meristematic tissues, green chloroplasts, red or yellow chromoplasts, the colorless plastids amylo- plasts and leucoplasts (starch containing), and elaioplasts (oil containing). The plastome is maternally inherited in most species of agricultural interest (Daniell, 2002; Hagemann, 2004). In maternal inheritance systems, paternal trans- mission of plastids is impeded during either the first pollen mitosis via unequal plastid distribution, or during generative or sperm cell development via plastid degeneration (Birky, 2001). Therefore, the generative and sperm cells in mature pollen tend to be free of plastids. Confinement of transgenic plastids in maternal tissues abrogates the ability of recombinant sequences to disperse to weedy relatives, or nearby agricultural stands. An additional level of control is offered by the recently developed system for reversible male sterility via light regulated expression of the phaA gene of Acinetobacter sp. Encoding -ketothiolase in the chloroplast genome (Ruiz and Daniell, 2005). 2. THE CHLOROPLAST GENOME 2.1. The Molecule and Its Genetic Transformation Plastid genome organization and structural features are conserved among eukaryotic photosynthetic organisms (Sugiura, 1992; Raubeson and Jansen, 2005). The circular molecule (figure 1) can be divided into three distinct domains: large single copy (LSC), small single copy (SSC) and the inverted repeat (IR) which is present in  PLASTID PATHWAYS 83 Figure 1. The Chloroplast Genome: Schematic representation of the plastid genomes of two Solanaceae species, Solanum tuberosum (potatoe) and Lycopersicion esculentum (tomatoe) exact duplicate separated by the two single copy regions. Restriction fragment length polymorphism (RFLP) analysis indicates that the molecule exists in two orientations present in equimolar proportions within a single plant (Palmer, 1983). The circular molecule undergoes interconversion to a dumbbell-shaped confor- mation that is believed to be facilitated by the presence of the IR. Concerted evolution within the IR (Kolodner and Tewari, 1975; Kolodner et al., 1976) suggests intramolecular recombination between the repeats is a possible mechanism. The plastid RecA homolog has demonstrated DNA strand transfer activity (Cerutti and Jagendorf, 1993) and is thought to be responsible for the site specific integration of foreign DNA sequences in the plastid genome by homologous recombination. Through successive rounds of regeneration on selective media the iteration of the transplastome is favored as plastids carrying the resistance marker, and in turn the 84 RUHLMAN AND DANIELL cells that harbor these plastids, are preferentially maintained as plastome molecules are divided up between daughter chloroplasts and subsequently as plastids are partitioned between daughter cells at mitosis (Moller and Moller, 2005). Over time all plastid genome copies, regardless of plastid type, carry the transplastome to the exclusion of detectable wild type copies, a condition referred to as homoplasmy. 2.2. Operons in Plastids Many chloroplast genes of crop plants are co-transcribed from operons producing polycistronic primary transcripts. Several monocistrons are also transcribed including psbA and rbcL, encoding the D1 core polypeptide of photosystem II, and Rubisco large subunit respectively, as well as most of the 30 tRNA genes (Bonen et al., 2004). 2.2.1. Translational regulation Transcript availability is not the rate limiting factor for protein accumulation in mature leaf chloroplasts. Transcript stability, maturation and subsequent translation into protein products are influenced by several features of the mRNA. Chloroplast mRNAs include 5’ and 3’ untranslated regions (UTRs), both of which confer on the molecule distinctive elements necessary for the eventual production of chloroplast proteins from mono- or polycistronic transcription units (Monde et al., 2000). Accumulation of polycistronic mRNAs as well as their efficient translation leading to high levels of foreign protein has been established (Jeong et al., 2004; Quesada- Vargas et al., 2005). Native and heterologous elements have been successfully employed for the regulation of foreign protein expression in chloroplast transformation experiments. The ability to drive protein accumulation in the plastid system has lead to the implementation of the psbA 5’ UTR in many transformation experiments where high levels of foreign product is desired (De Cosa et al., 2001; Fernandez-San Milan et al., 2003; Leelavathi and Reddy 2003; Dhingra et al., 2004; Molina et al., 2004; Watson et al., 2004; Devine 2005; Chebolu and Daniell 2007; Viitanen et al., 2004). Due to its dependence on light for activation of translation this element has an obvious limitation in non green tissues. An alternative to this highly utilized sequence that has been employed successfully for expression in green (Guda et al., 2000; Staub et al., 2000) and nongreen (Kumar et al., 2004a) tissues is the 5’ translation control region of bacteriophage T7 gene 10 (T7g10). In addition to its ability to associate with plastid ribosomes in a light-independent manner this sequence element should be free of developmental regulation in plant plastids. Utilization of heterologous translational elements may be advan- tageous in several regards. Chloroplast transformation for multi-gene constructs will require the implementation non-repetitive translation sequences for transgenes to avoid looping out events leading to the elimination of inserted sequences. Furthermore foreign signals should not detract from or have to compete with the  PLASTID PATHWAYS 85 expression of endogenous proteins. This could be a benefit for foreign protein accumulation and possibly deter possible interference with the expression of native proteins. 3. PLASTID OPERONS AND MULTIGENE ENGINEERING That the plastid expression system allows for the transcription of operons from a single promoter to produce translatable polycistronic mRNAs (De Cosa et al., 2001; Ruiz et al., 2003; Quesada-Vargas et al., 2005) offers great potential for metabolic engineering. The ability to transform the plastome for multiple genes in a single recombination event makes possible the expression of multi-enzyme pathways in the first transformed generation eliminating the need to cross lines recombinant for individual genes. Simultaneous integration of selectable markers along with genes of interest assures that regenerants expressing said markers will harbor the entire transformation cassette. Multi-gene engineering permitted the ‘double gene single selection’ system (Kumar et al., 2004b) that facilitated the generation of homoplasmic cotton transformants through the ability to apply selective pressure in green as well as non-green stages of development. 3.1. Chaperones While the plastid is less protease rich than the cytoplasm, it is not free from proteolytic activity (Adam and Clarke, 2002). Inclusion of the native chaperonin of the Bacillus thuringiensis (Bt) cry2Aa2 operon led to an abundance (about 46% of TSP) of Cry protein in transplastomic tobacco plants (De Cosa et al., 2001), the highest level of transgene expression reported for plants. An upstream sequence of the operon, orf 2, encodes a protein involved in Cry folding into cuboidal crystals. In its crystalline form Cry is highly resistant to proteolytic activity (Crickmore and Ellar, 1992; Staples et al., 2001). As co-integration of such folding factors is feasible through plastid transformation, a worthwhile approach would be to identify native or general factors for proteins of interest which are susceptible to degradation. Operonal expression conferred high levels of tolerance to the organomercuial compound phenylmertcuric acetate (PMA) on chloroplast transformants for the merA and merB genes. Originating from bacteria, mercuric ion reductase and organomercurial lyase are encoded by merA and merB respectively. These enzymes convert toxic methyl-mercury (CH2 Hg) to elemental mercury (Hg(0)). The latter is much less toxic, readily volatilized and should theoretically be released by plants to the atmosphere through transpiration (Rugh et al., 1996; Bizily et al., 1999; Bizily et al., 2000). Previous studies employing nuclear transformation and subse- quent breeding to establish both genes in transgenic Arabidopsis demonstrated that expression of both genes was required to confer significant resistance to toxic mercury (Bizily et al., 1999; Bizily et al., 2000). Measurements of Hg(0) evolution correlated well with the ability of these plants to survive on the highest concen- tration reported in this work: 5 μM of PMA. Chloroplast transformants expressing 86 RUHLMAN AND DANIELL the native bacterial sequence as an operon with SD-type signal driving translation were not susceptible to 100 μM of PMA and were able to survive at concentrations up to 400 M of PMA (Ruiz et al., 2003). 3.2. Vitamins 3.2.1. Vitamin A Improvement of the nutritive value of seeds and vegetative tissues has been a goal of genetic engineers for some time. In Southeast Asia, it is estimated that up to half a million children go blind each year because of vitamin A deficiency (VAD). An estimated 140–250 million children suffer subclinical VAD which exacerbates afflictions such as diarrhea, respiratory ailments, and other childhood diseases such as measles. Supplementation with vitamin A reduces mortality from measles by 50% (WHO). As rice is the predominant staple in Southeast Asia it has been a primary target for engineering of vitamin A. The development of a nuclear transgenic rice lines expressing three separate genes involved in provitamin A synthesis: phytoene synthase, phytoene desaturase, -carotene desaturase (the two latter activities, separate in plants, were mediated by the single bacterial enzyme CrtI) and lycopene -cyclase, has come as the result of many years of effort and required transformation with three different vectors (Ye et al., 2000, Datta et al., 2003; Baisakh et al., 2006). The precursor for -carotene, geranylgeranyl-diphosphate, is synthesized in wild type rice endosperm plastids, the site to which the enzymes were targeted. The ability to introduce several genes in a single transformation event through plastid genetic engineering could be used to express metabolic pathways, such as provitamin A synthesis, directly in plastids. 3.2.2. Vitamin E The world vitamin E market was valued at more than one billion dollars in 1995 with synthetically produced comprising approximately 88% of the supply (Herbers, 2003). This essential nutrient is thought to provide protection from some cancers due to its antioxidant properties (Maeda et al., 1992; Pastori et al., 1998; Santillo and Lowe, 2006; Shiau et al., 2006). The tocopherols and tocotrienols that comprise the group of lipid species collectively known as vitamin E are formed and accumulate in the photosynthetic plastids and cyannobacteria. Several experi- ments employing nuclear transformation to overexpress various enzymes involved in vitamin E synthesis have been carried out. These enzymes are nuclear encoded and imported to plastids, with the exception of 4-hydroxyphenylpyruvate dioxy- genase (HPD) which is active in the cytoplasm catalyzing a single step in the pathway leading to tocopherol. Efforts to increase substrate availability through expression of plastid localized, feedback insensitive enzymes (Falk et al., 2005) and overexpression of homogen- tisate phytyltransferase (HPT) (Collakova and DellaPenna, 2003a, 2003b), whose  PLASTID PATHWAYS 87 activity represents the first committed step in tocopherol synthesis, have demon- strated the possibility to improve vitamin E availability in plants. In a recent study the HPD enzyme from barley was expressed in tobacco chloroplasts. These transplastomic plants accumulated more than twice as much -tocopherol in leaves than wild type plants, while no significant difference was detected in seeds (Falk et al., 2005). The authors speculate that HPD product HGA must interact with HPT through a membrane interaction and that HGA produced in the plastid is restricted in this association. This is a reasonable assumption as in the wild type HGA migrates back across the plastid envelope for continuation of the pathway. Unfortunately the tobacco plants in these experiments are not knocked out for endogenous HPD activity and therefore a definitive conclusion cannot be drawn from these results. Investigation with nuclear transformants overexpressing HPT has demonstrated that this activity is limiting in the pathway leading to vitamin E (Collakova and DellaPenna, 2003a). This suggests that simply flooding HPT with substrate may not necessarily lead to an enhancement of tocopherol synthesis. In a recent review on metabolic engineering for nutritional enhancement of plants, it was proposed that at least five genes in the vitamin E pathway may have to be upregulated in some way to significantly enhance its accumulation in an oilseed crop (Kinney, 2006). Transformation of the plant chloroplast with any or all of these genes via multi-gene constructs would allow their expression to be controlled at several levels. All of these genes could be transcribed as an operon from a single promoter, with tailoring of translation made possible through the use of UTRs specific for develop- mental stages or light environment. Especially promising may be the use of transplas- tomic soybean in future studies as plastid expression is prolonged in the mature seed (Dufourmantel et al., 2005) compared to other non-photosynthetic oil seeds. 3.3. Amino Acids In addition to vitamins, several amino acids are at least partially synthesized in the plant chloroplast by enzymes which are nuclear encoded (see Table 1), providing an insight into how we may better approach nutritional enhancement of plants. The ability to transfer groups of genes, for example enzymatic pathways, in single transformation construct makes the chloroplast an attractive site for the engineered expression of essential amino acids. Methionine, lysine, threonine and isoleucine constitute the aspartate family of amino acids. With the exception of methionine these amino acids are synthe- sized entirely in the plastid from a parent molecule, aspartic acid, imported from the cytosol. Methionine synthesis is completed by enzymes localized outside the plastid. The biosynthesis pathways of methionine and threonine diverge from O-phosphohomoserine (OPHS), an intermediate metabolite in the aspartate family of amino acids (Figure 2). OPHS represents the common substrate for both Threonine Synthase (TS) and Cystathionine--Synthase (CGS). OPHS is either directly converted to threonine by TS, or, in a three-step mechanism, to methionine through condensation of cysteine and OPHS to cystathionine, which is subsequently 88 RUHLMAN AND DANIELL Table 1. Abbreviated list of nuclear-encoded metabolic enzymes post-translationally imported into plastids. Adapted from http://www.chloroplast.net Amino Acid Metabolism Aspartate Kinase (EC 2.7.2.4) Lys and Thr; involved in generation of homoserine for Met, Gly, Ser metabolism Homoserine Dehydrogenase (EC 1.1.1.3) Threonine, Isoleucine and Methionine Aspartate Semialdehyde Dehydrogenase produces branchpoint intermediate to Lys and (EC 1.2.1.11) Thr or Met synthesis Diaminopimelate Decarboxylase (EC 4.1.1.20) final enzyme in Lysine synthesis Diaminopimelate Epimerase (EC 5.1.1.7) Lysine synthesis Dihydrodipicolinate Reductase (EC 1.3.1.26) Lysine; formation of tetrahydrodipiolinate Dihydrodipicolinate Synthase (EC 4.2.1.52) first reaction unique to Lys synthesis; senitive to Lys feedback inhibition Homoserine Kinase (EC 2.7.1.39) forms branchpoint intermediate in pathways for Met and Thr Succinyldiaminopimelate Transaminase Lysine (putative) (EC 2.6.1.17) Threonine Synthase (EC 4.2.3.1) first committed step in Thr synthesis 2-isopropylmalate Synthase (EC 2.3.3.13) introductory enzyme in Leu synthesis 3-isopropylmalate Dehydratase, large and small branched chain amino acid synthesis subunit (EC 4.2.1.33) 3-isopropylmalate Dehydrogenase (EC 1.1.1.85) Leucine synthesis Acetolactate Reductoisomerase (ketol-acid reduces acetohydroxyacids to dihydroxyacids reductoisomerase) (EC 1.1.1.86) (branched chain pathway) Acetolactate Synthase (Acetohydroxy Acid entry point to branched chain pathway; Synthase) catalytic and regulatory subunit herbicide target (ALS inhibitors) (EC 2.2.1.6) Branched-chain amino acid Aminotransferase Valine, Leucine and Isoleucine (last in (EC 2.6.1.42) synthesis, first in degredation) Dihydroxy Acid Dehydratase (EC 4.2.1.9) branched chain pathway, third enzyme Threonine Deaminase (Threonine Dehydratase) Ile synthesis; conversion of L-threonine to (EC 4.3.1.19) -ketobutyrate 3-phosphoglycerate Dehydrogenase (EC 1.1.1.95) Glycine and Serine; entry point to Ser, Gly pathway 3-phosphoserine Phosphatase (EC 3.1.3.3) 3-phosphoserine to serine: final step Phosphoserine Aminotransferase (EC 2.6.1.52) Serine 3-dehydroquinate Dehydratase (EC 4.2.1.10)/ Tryptophan, Tyrosine and Phenylalanine Shikimate 5-dehydrogenase (EC 1.1.1.25) (bifunctional enzyme) 3-dehydroquinate Synthase (EC 4.2.3.4) pre-chorismate pathway; aromatic amino acids 3-deoxy-7-phosphoheptulonate Synthase first enzyme of the shikimate pathway (EC 2.5.1.54) Shikimate Kinase (EC 2.7.1.71) phosporylation of shikimate; aromatic amino acid synthesis 5-enolpyruvylshikimate-3-phosphate (EPSP) target of glyphosate herbicides synthase (EC 2.5.1.19) Chorismate Synthase (EC 4.2.3.5) product chorismate is last common precursor to numerous aromatic compounds Anthranilate Phosphoribosyltransferase Tryptophan biosynthesis (EC 2.4.2.18) Anthranilate Synthase - and -subunits chorismate to anthranilate (EC 4.1.3.27) Pretyrosine (arogenate) Dehydrogenase Tyrosine biosynthesis (EC 1.3.1.43)  PLASTID PATHWAYS 89 Chorismate Mutase (EC 5.4.99.5) shikimate pathway, first enzyme leading to Tyr and Phe Indole-3-glycerol-phosphate Synthase Tryptophan (EC 4.1.1.48) Phosphoribosylanthranilate Isomerase Tryptophan (EC 5.3.1.24) Prephenate Dehydratase (EC 4.2.1.51) Phenylalanine biosynthesis; prephenate to phenylpyruvate Tryptophan synthase - and -subunits final step in Tryptophan synthesis (EC 4.2.1.20) Glutamate-cysteine Ligase (EC 6.3.2.2) first step in glutathione synthesis Glutathione Synthetase (EC 6.3.2.3) Cysteine/sulfur metabolism; Glutathione synthesis ATP Sulfurylase (sulfate adenylyltransferase) Cysteine and Methionine; sulfur assimilation (EC 2.7.7.4) Cystathionine beta lyase (EC 4.4.1.8) Cysteine and Methionine; sulfur metabolism (cystathionine to homocysteine) Cystathionine gamma Synthase (EC 2.5.1.48) Cysteine and Methionine; sulfur metabolism (cystathionine from homocysteine) O-acetylserine (thiol)-lyase (OAS TL) Cysteine and Methionine; sulfur metabolism (EC 2.5.1.47) (Cysteine Synthase complex) Serine O-acetyltransferase (EC 2.3.1.30) Cysteine and Methionine; sulfur metabolism (Cysteine Synthase complex) Acetylornithine Deacetylase (EC 3.5.1.16) Arginine metabolism; last step from glutamate to ornithine (catabolism) Acetyl Ornithine Transaminase Arginine; glutamate to ornithine (catabolism) (EC 2.6.1.11) Delta-1-pyrroline-5-carboxylate Reductase final step in Proline synthesis (EC 1.5.1.2) N-acetylglutamate Kinase (EC 2.7.2.8) key regulatory step in Arginine synthesis N-acetylglutamate Synthase Ornithine synthesis (EC 2.3.1.1)/N-acetylornithine Glutamate Acetyltransferase (EC 2.3.1.35) Ornithine Carbamoyltransferase (EC 2.1.3.3) Arginine synthesis Carbamoylphosphate Synthetase A and B Pyrimidine and Arginine metabolism subunits (EC 6.3.5.5) 1-(5-phosphoribosyl)-5-[(5- Histidine synthesis phosphoribosylamino)methylideneamino] imidazole-4-carboxamide isomerase (EC 5.3.1.16) ATP-phosphoribosyl Transferase (EC 2.4.2.17) first step in Histidine synthesis Glutamine Amidotransferase (EC 2.4.2.-) / Histidine synthesis Cyclase (EC 4.1.3.-) Imidazoleglycerol-phosphate Dehydratase Histidine synthesis; target for the triazole (EC 4.2.1.19) phosphonate herbicides Phosphoribosyl-ATP pyrophosphohydrolase Histidine synthesis (EC 3.6.1.31) / phosphoribosyl-AMP cyclohydrolase (EC 3.5.4.19) Fatty Acid Metabolism Glutathione Peroxidase Phospholipid sulfur metabolism; development and stress Pydroperoxide (EC 1.11.1.12) (EC 1.11.1.9) response Acetyl-CoA Carboxylase (EC 6.4.1.2) Acetyl-CoA, malonyl-CoA synthesis (continued ) 90 RUHLMAN AND DANIELL Table 1. (continued ) Fatty Acid Metabolism heteromeric form Biotin carboxyl carrier protein Carboxyltransferase alpha subunit Biotin Carboxylase (EC 6.3.4.14) subunit Acetyl-CoA Synthetase (EC 6.2.1.1) [acyl carrier protein] S-malonyltransferase Elongase (EC 2.3.1.39) Pyruvate Dehydrogenase  and  subunits (EC 1.2.4.1) Dihydrolipoamide S-acetyltransferase Pyruvate Dehydrogenase subunit (EC 2.3.1.12) 1-acylglycerol-3-phosphate O-acyltransferase synthesis of phosphatidic acid /lysophosphatidic acid acyltransferase (EC 2.3.1.51) Chlorophyll Synthesis Chlorophyll Synthesis Glutamyl-tRNA Synthetase (EC 6.1.1.17) L-glutamate to glutamate 1-semialdehyde (Mg-tetrapyrrol synthesis); amino acid synthesis Aminolevulinate Dehydratase (EC 4.2.1.24) chlorophyll synthesis Chlorophyll a Oxygenase conversion of a methyl group to a formyl side group (chlorophyll b synthesis) Chlorophyll Synthetase (EC 2.5.1.62) chlorophyll a synthesis; attachment of phytyl side chain Coproporphyrinogen III Oxidase (EC 1.3.3.3) formation of protoporphyrin IX Glutamate-1-semialdehyde Aminotransferase formation of delta-aminolevulinate from (EC 5.4.3.8) glutamate 1-semialdehyde Magnesium Chelatase, Chl D, H, I subunits branchpoint in chlorophyll/heme synthesis (EC 4.99.1.-) Magnesium Protoporphyrin IX Methyltranserase chlorophyll metabolism; Mg-dependent pathway (EC 2.1.1.11) Protochlorophyllide Reductase (EC 1.3.1.33) (POR) conversion to chlorophyllide subunits A, B and C Uroporphyrinogen III Synthase (EC 4.2.1.75) chlorophyll metabolism Carotenoid and Xanthopyll Synthesis −carotene epsilon-hydroxylase lutein synthesis −carotene Hydroxylase (EC 1.14.13.-) zeaxanthan synthesis Lycopene epsilon-cyclase -and -carotene synthesis 9-cis-epoxycarotenoid Dioxygenase (neoxanthin carotenoids to ABA through xanthoxin cleavage enzyme) Carotenoid Isomerase lycopene synthesis Lycopene -cyclase (EC 1.14.-.-) -and -carotene synthesis Phytoene Desaturase (EC 1.3.99.-)/Phytofluene pytoene to -carotene Desaturase Phytoene Synthase (EC 2.5.1.32) first committed step in carotenoid synthesis Violaxanthin De-epoxidase violaxanthin to zeaxanthin Zeaxanthin Epoxidase zeaxanthin to violaxanthin, entrance to ABA biosynthetic pathway −carotene Desaturase (EC 1.14.99.30) lycopene synthesis Carotenoid Associated Protein  PLASTID PATHWAYS 91 Pigments, Carotenoid and Plastoquinone 1-deoxy-D-xylulose 5-phosphate MVA- independent pathway; synthesis of C5 Reductoisomerase (EC 1.1.1.267) isoprenoid intermediates isopentyl diphosphate, dimethylallyl diphosphate 2C-methyl-D-erythritol 2,4-cyclodiphosphate Geranylgeranyl diphosphate synthesis Synthase (EC 4.6.1.12) 4-diphosphocytidyl-2C-methyl-D-erythritol Geranylgeranyl diphosphate synthesis Kinase (EC 2.7.1.148) 4-diphosphocytidyl-2C-methyl-D-erythritol Geranylgeranyl diphosphate synthesis Synthase (EC 2.7.7.60) 4-hydroxy-3-methylbut-2-en-1-yl diphosphate Geranylgeranyl diphosphate synthesis Synthase (EC 1.17.4.3) 4-hydroxy-3-methylbut-2-enyl diphosphate Geranylgeranyl diphosphate synthesis Reductase (EC 1.17.1.2) Geranyl Diphosphate Synthase (EC 2.5.1.1) short chain prenyltransferase Geranylgeranyl Diphosphate Synthase branchpoint to gibberillins, carotenes, (EC 2.5.1.29) vitamin K2 and E, chloropyll, tyrosine, ABA Isopentenyl Diphosphate Isomerase Diverts isopentenyl diphosphate to cytokinin pthwy (EC 5.3.3.2) Monoterpene Sesquiterpene Synthase like Isoprenoid synthesis; terpenes including TAXOL/ proteins including taxa-4(5),11(12)-diene PACLITAXIL Plastidal entry point to synthase, a diterpene cyclase gibberellin pathway 3-oxo-5-alpha-steroid 4-dehydrogenase carotenoids to brassinolide (via squalene) (EC 1.3.99.-) Cinnamate 4-hydroxylase (EC 1.14.13.11) anthocyanin pathway; flavanoid synthesis Phenylalanine Ammonia Lyase (EC 4.3.1.5) anthocyanin pathway; flavanoid synthesis; first enzyme Secondary Metabolism Adenine Phosphoribosyltransferase (EC 2.4.2.7) caffeine synthesis; purine nucleotide synthesis Adenosine Kinase (EC 2.7.1.20) caffeine synthesis; purine nucleotide synthesis Groporphyrin III Methylase (EC 2.1.1.107) Siroheme synthesis Vitamins 4-methyl-5(b-hydroxyethyl)-thiazole Thiamine synthesis (vitamin B1) Monophosphate Biosynthesis Protein Hydroxyethylthiazole Kinase (EC:2.7.1.50) Vitamin B1 Phosphomethylpyrimidine Kinase Vitamin B1 (EC 2.7.4.7)/Thiamine-phosphate Pyrophosphorylase (EC 2.5.1.3) Thiamin Biosynthesis Protein Vitamin B1 Geranylgeranyl Reductase branchpoint to vitamins K2 and E, and chlorophyll Homogentisate Phytyltransferase Vitamin E - -Tocopherol Tocopherol Cyclase Vitamin E - -Tocopherol Gamma Tocopherol Methyltransferase Vitamin E - -Tocopherol (EC 2.1.1.95) 1.4-dihydroxy-2-naphthoate (DHNA) Vitamin K2 – Phylloquinone Phytyltransferase (EC 2.5.1.-) 2-Oxoglutarate Decarboxylase (EC 4.1.1.71) / Vitamin K2 – Phylloquinone SHCHC synthase (EC 4.1.3.-) C-methyltransferase (EC 2.1.1.-) vitamin K2 – Phylloquinone; putative ubiE homologue Isochorismate Synthase (EC 5.4.99.6) shikimate pathway O-succinylbenzoate Co-A Ligase (EC 6.2.1.26) Vitamin K2 – Phylloquinone 92 RUHLMAN AND DANIELL converted to homocysteine and then methionine. Thus, the enzymes CGS in the methionine pathway and TS in the threonine pathway compete for the common substrate, OPHS (Avraham and Amir, 2005). The fact that plant TS and CGS are branch point enzymes competing for the same substrate demands the effective regulation of the respective enzymatic activities. 3.3.1. Enhancing Sulfur amino acid accumulation Over-expression of CGS in potato increased methionine levels in the leaves, roots and tubers 6 fold (Di et al., 2003). Antisense of TS in potato transgenic lines reduced threonine to 45% compared to wild-type controls, whereas methionine levels increased up to 239-fold depending on the transgenic line and environmental conditions (Zeh et al., 2001). Increased levels of homoserine and homocysteine indicate increased carbon allocation into the aspartate pathway. Tubers of TS antisense potato plants contained a methionine level increased by a factor of 30 but showed reduction in threonine. These plants offer a major biotechnological advance toward the development of crop plants with improved nutritional quality. Methionine is considered the most limiting amino acid and efforts to improve the sulfur content of seed and forage crops has fostered the body of work exploring ways to achieve this through the accumulation of methionine and cysteine in vegetative tissues and storage organs. Sulfur-rich seed storage proteins such as the 2S albumins from sunflower seed (SSA), Brazil nut, and pea vicilin have been employed in Figure 2. Amino Acid Biosynthesis in Chloroplasts: Many metabolic pathways, including several for the synthesis of essential amino acids are, localized to plastids. Black lines indicate feedback regulation (+ or −). Dashed lines indicate this regulation has not been observed in all species investigated, modified from http://www.compulink.co.uk/∼argus/Dreambio/photosynthesis/chloroplast 4.gif  PLASTID PATHWAYS 93 nuclear transformation experiments. Initial experiments with the Brazil nut albumin resulted in detectable expression in all tissues and organs examined due to the inclusion of a constitutive promoter. The gene product was localized predominantly to mesophyll vacuoles where it could not accumulate to high levels (Muntz et al., 1998). The vegetative vacuole is a protease rich compartment and this may have led to rapid degradation of the foreign protein. In subsequent work with SSA and pea vicilin, the fusion of a C-terminal endoplasmic reticulum (ER) targeting signal (KDEL) has facilitated varying degrees of accumulation in nuclear transgenic plants. In tobacco plants expressing the pea vicilin gene addition of the KDEL sequence resulted in a 100-fold enhancement of this recombinant protein and extended its half life by a factor of 10 (Wandelt et al., 1992). For both tobacco and alfalfa, ER targeting of SSA produced “easily detectable” expression of foreign protein in western analysis as compared to transformants for the SSA gene alone in which no SSA was detected (Tabe et al., 1995), again confirming the contrasting results observed in leaves & seeds. Similar results were found for SSA expression in Trifolium subterraneum L. (Khan et al., 1996). It has been suggested that the availability of soluble methionine limits the potential accumulation of SSA in nuclear transgenic lines. Experiments have been carried out with the methionine-rich zein genes of maize to address the same limitation for its accumulation in alfalfa (Bagga et al., 2004; Amira et al., 2005) and tobacco (Amira et al., 2005). Transgenic alfalfa and tobacco lines expressing ß-zein were crossed with those expressing Arabidopsis cystathionine--synthase (atCGS). Soluble methionine levels were reduced when compared to plants expressing atCGS alone in both species and ß-zein expression was significantly enhanced in alfalfa. In tobacco leaf samples from plants co-expressing atCGS and ß-zein, significant accumulation of ß-zein was not observed compared to those expressing ß-zein alone, suggesting that cytosolic translation of this mRNA is not limited by methionine availability. Proteins with high sulfur content are likely to be stabilized by disulfide bridges. Improved accumulation of sulfur amino acids in ER targeting experiments suggests this may be the case. This compartment contains functional PDI for the estab- lishment of such stabilizing bonds (Denecke et al., 1992; Levitan et al., 2005). Furthermore those products targeted to the ER would not be subject to the protease- rich environment of the cytoplasm. Chloroplast may prove to be an ideal site for the expression of sulfur rich seed storage proteins as PDI is active in this compartment (Trebitsh et al., 2001; Kim and Mayfield, 2002; Levitan et al., 2005; Alergand et al., 2006). In addition export of proteins translated on plastid ribosomes across the organelle’s double membrane has not been reported suggesting they are retained within the chloroplast. 3.3.2. Improvement of lysine and threonine content Cereal grains represent a major constituent in the diet of humans and livestock world wide. Lysine and threonine are limiting amino acids in cereal grains and improving the lysine content of fodder for poultry and swine has received consid- erable attention. Lysine metabolism in plants is regulated primarily by feed back 94 RUHLMAN AND DANIELL inhibition of plastid localized dihydrodipicolinate synthase (DHPS). To address this limitation on lysine accumulation, the feed back insensitive bacterial DHPS has been transferred to the nuclear genome of tobacco (Shaul and Galili, 1993; Karchi et al., 1994), soybean and canola (Falco et al., 1995; Mazur et al., 1999). Constitutive expression facilitated significant accumulation of free lysine in tobacco leaves but not without serious deleterious effects (Perl et al., 1992). Expression from a seed specific promoter provided amelioration of this effect at the expense of lysine production (Karchi et al., 1994). Moderate accumulation of free lysine was observed in developing seeds of transgenic tobacco via the expression of two bacterial enzymes, DHPS and feedback insensitive Aspartate Kinase but content was not significantly higher than the wild type when seeds reached maturity. It was found that the presence elevated levels of free lysine stimulated the catabolic activity of the cytosolic enzyme Lysine-Ketoglutarate Reductase (Karchi et al., 1994; Tang et al., 1997; Zhu et al., 2001). Subsequent seed specific expression of the bacterial DHPS in soybean and canola demonstrated a significant enrichment of lysine over wild type seeds. Unfortunately this resulted in problems with seed germination and the presence of various catabolites of lysine was noted (Falco et al., 1995). A more successful approach for lysine accumulation, particularly in cereal grains, has been transformation with genes encoding proteins with a high proportion of this amino acid in its sequence. Several genes have been employed in maize to relative success. Constructs incorporating the barley high lysine (BHL) or hordothionine led to an elevation of lysine content in transgenic seed (in Galili et al., 2005). Soybean vegetative storage protein -subunit (S-VSP), also rich in lysine, was co-expressed constitutively with the bacterial DHPS in transgenic tobacco. S-VSP accumulation was higher in these plants than in those expressing the DHPS alone with total lysine level increased by 30% over wild type (Guenoune et al., 2003). Designing experiments to include both the enzymatic source of desired amino acids as well as a proteolytically stable “sink” polypeptide or polymer may be the route to their stable accumulation in plant tissues. Chloroplast transformation may support this end in several ways. Multi-gene engineering, various benefits facilitated by compartmentalization (i.e. relief of co suppression, protection from potential degredative influences in the cytoplasm) and the presence of plastid PDIs are among the features that make plastid biotechnology a worthwhile approach for explorations such as those described herein. 4. CHLOROPLASTS AS BIOREACTORS: CAN WE REPLACE PLANTS WITH PLANTS? Technology has brought our society a long way in terms of our ability to produce desirable products on a large scale in industrial plant operations. The mandate for future productivity includes the development of technologies that meet the demands of a growing population without concomitant threat to the environment. In the plastids of green plants we are finding the potential to express desired protein products, and, through the engineering of biosynthetic pathways, accumulate important non-protein molecules in a cost-effective manner.  PLASTID PATHWAYS 95 4.1. Pharmaceutical Products A number of pharmaceutical proteins have been successfully expressed in trans- genic plastids including various therapeutic agents and vaccine antigens (Daniell et al., 2001a,b; Fernandez San-Millan et al., 2003; Tregoning et al., 2004; Watson et al., 2004; Koya et al., 2005). Expression of foreign proteins as fusion partners has facilitated significant accumulation of human somatotropin (hST) and interferon gamma (IFN-g) in transplastomic tobacco. While plastid transformants expressing IFN-g alone were found to accumulate 100 times as much of the foreign protein than nuclear transformants for the same gene, up to 0.1% of TSP (Leelavathi and Reddy, 2003), for feasible purification of therapeutic proteins from plant tissues accumulation will have to reach approximately 1% of TSP (Daniell 2006). Transfor- mation of tobacco plastids with a new construct which carried the ifnG gene fused to the uidA gene for -glucaronadase (GUS), a much more stable protein, allowed the IFN-g fusion to accumulate to an estimated level of 6% of TSP (Leelavathi and Reddy, 2003). Previously, the fusion approach had facilitated the accumulation of human somatotropin (hST). Initial expression of hST reached up to 0.2% of TSP in plastid transformants but when expressed as a fusion with Ubiquitin, hST accumulation in tobacco leaf tissues reached up to 7% of TSP, more than 300 times the level seen in nuclear transformants for the same gene (Staub et al., 2000). Although plastids are not thought to encode any disulfide bonded proteins, many nuclear encoded examples are ultimately expressed within the organelle. Plastid- localized PDI catalyzes disulfide bond formation for imported proteins and its presence may have contributed to the increased stability of hST in plastids. Human Serum Albumin (HSA) is the major component in blood and is used to replace blood volume in many clinical and trauma situations. In its mature form the multimeric globular protein is stabilized by 17 disulfide bonds. Potato nuclear transformants in which HSA expression was targeted to the tuber apoplast have been reported to accumulate up to ten times more of the foreign protein than transgenic plants expressing HSA in the cytoplasm ( Farran et al., 2002). Expression in the protease- poor apoplastic space can provide protection for foreign proteins susceptible to proteolytic degradation. Where expression of HSA was driven by psbA 5’UTR transplastomic plants accumulated more than 50 times as much protein as nuclear tranasforments localizing HSA to the apoplast (Fernandez-San Milan et al., 2003). The ability to carry out post-translational modifications, such as disulfide bond formation, is likely a major contributing factor in limiting HSA degredation by proteases in the plastid. Abundant local expression of HSA in tobacco plastids lead to the formation of inclusion bodies which allowed for a facile purification strategy yielding a recovery of 0.25 mg HSA/g fresh weight, well within the range of industrial-scale feasibility. In addition to added stability, correct disulfide bond formation in some therapeutic proteins is an absolute requirement for functional, biologically active molecules. For example, the CTB/GM1 system described above relies on the ability of CTB monomers to assemble in the pentameric form for GM1 binding and subsequent internalization. Each monomer contains an intramolecular disulfide bridge which is 96 RUHLMAN AND DANIELL essential to subunit association (Ludwig et al., 1985; Dertzbaugh and Cox, 1998). In vitro GM1 binding and in vivo epithelial absorption following oral delivery of plastid-derived CTB fusion proteins, such as CTB-GFP and CTB-Pins, demonstrates that the integrity of the pentamer is preserved in minimally processed plant tissues (Limaye et al., 2006; Ruhlman et al., 2007). One of the limitations to the production of affordable therapeutic proteins by bacterial fermentation is the requirement for post-harvest modification to achieve correct structural conformation. Treatment of hepatitis C with interferon-2b (IFN-2b), currently produced in E. coli, is estimated to cost up to $26,000 per year. Biologically active INF-2b contains two disulfide bonds and complete loss of antiviral activity results from their reduction (Morehead et al., 1984; Bodo and Maurer-Fogy, 1986). When expressed in tobacco plastids INF-2b accumulated up to 20% of tsp, or 3 mg g−1 of fresh leaf weight. In vitro and in vivo assays using crude transplastomic leaf extracts and purified plastid-derived INF-2b demon- strated biological activity comparable to the commercial product produced in E. coli (Arlen et al., 2007). The possibility to introduce several genes as a single construct allows for the expression of multisubunit complexes with transcription of said genes under the control of a single promoter. Having a readily available mRNA pool may help to maximize the potential for correct association into functional units and fine tuning for stoichiometry could be accomplished through manipulation of 5’ translation elements. Operon engineering in plastids has been employed to express the heavy (H) and light (L) chain of the Guy’s 13 antibody. This human monoclonal antibody recognizes streptococcal antigen I/II (SA I/II), a major cell surface glycoprotein of Streptococcus mutans. In clinical trials this antibody has been shown to prevent colonization in the human oral cavity. The H and L chain were expressed in plastids where they were shown to assemble to the 160 kDa antibody by non- reducing western analysis with no concomitant deleterious effects observed in the transformants. The H and L chains associated through disulfide bridges formed by the plastid apparatus (Daniell et al., 2005b). In nuclear transformants, after several reproductive crosses to achieve co-expression, inclusion of the J chain and the secretory component along with H and L led to the accumulation of total assembled IgA/G of 5% of TSP (Ma et al., 1995; Ma et al., 1998). This result suggests chloroplast operonal expression for secretory antibodies such as Guy’s 13 merits further investigation. 4.2. Biopolymer Production 4.2.1. Polyhydroxybutyrate The pleiotropic effects that have constrained cytoplasmic expression of certain foreign proteins have not been limited to enzymes involved in amino acid and vitamin synthesis. Nocent phenotypes have been observed in transformation scenarios aimed at the accumulation of engineered biopolymers (Nawrath et al., 1994; Wrobel et al., 2004). Nuclear transformation experiments have utilized the  PLASTID PATHWAYS 97 plastid as a storage vessel in such instances. Polyhydroxybutyrate (PHB) is synthe- sized from acetyl-coenzyme A by the consecutive activity of three enzymes of bacterial origin: -ketothiolase, acetoacetyl-CoA reductase, and PHB synthase. In the bacteria PHB serves as a carbon storage molecule but it has attracted consid- erable interest from industry due to its potential application in biodegradable plastics and elastic polymers. Crosses made of parental lines transformed for two of the three enzymes, phaB and phaC, in initial experiments with Arabidopsis (in Nawrath et al., 1994) constituted the first report of PHB production in plants. Despite low yield of PHB these plants presented severely stunted phenotypes. Subsequent exper- iments fused the constitutive 35S CMV promoter and the transit peptide sequence of the small subunit of pea ribulose-bisphosphate carboxylase to the genes for each of the three enzymes. A similar strategy of sexual crosses led to Arabidopsis plants expressing substantial amounts of PHB in chloroplasts (Nawrath et al., 1994). An alternative to this approach would be to introduce this pathway, in a single transformation event, directly into the plastid genome for localized expression. While initial attempts resulted in tandem integration of the operon, only very low abundance of PHB in the leaves of transplastomic tobacco plants (Nakashita et al., 2001). An improved vector carrying a multigene construct delivered to the Arabidopsis nuclear genome facilitated abundant accumulation of PHB in chloro- plasts, unfortunately with concomitant deleterious effects on plant fitness (Bohmert et al., 2000). Later experiments integrating the operon into the inverted repeat region of the tobacco plastome demonstrated enhanced PHB accumulation but phenotypic abnormalities including growth inhibition and sterility accompanied the improvement (Lossl et al., 2003). In order to address nocuous side effects of PHB and other foreign proteins in the chloroplast recent investigations have sought to establish an inducible system for control of transgene expression (Lossl et al., 2005; Muhlbauer and Koop, 2005). Among the many attractive aspects of chloroplast transformation technology is the level of transgene containment achieved through the maternal inheritance of plastids in most agronomically important species. In an effort to diagnose the impediment to effective expression of the phb operon in transplastomic tobacco Ruiz and Daniell (2005) have discovered a novel and reversible instrument for induction of cytoplasmic male sterility (CMS). CMS is commonly associated with incompatability between nuclear and mitochondrial genomes and has been addressed in the literature as a deterrent to transgene escape via pollen and the production of commercial F1 hybrid seed (Saeglitz et al., 2000; Havey, 2004; Chase, 2006). During the course of their investigation into the light regulated expression of -ketothiolase in plastid transformants none of the severe pleiotropic effects, such as chlorosis and retarded growth, which had been found in plants expressing the entire operon were observed. The persistent characteristic in these plants was the abberant morphology of the male reproductive structures including shortened filaments and pollen grains that were collapsed or absent altogether. This phenotype persisted in the T1 plants generated by fertilization with wild type pollen. It was hypothesized that the male sterile effect resulted from a depletion of the acetyl-CoA pool in 98 RUHLMAN AND DANIELL plastids expressing -ketothiolase. Restoration of male fertility was accomplished by continuous illumination of transplastomic plants. The reducing environment facilitiated by continuous exposure to light favored the activity of ACCase enabling it to compete with -ketothiolase for the available substrate, acetyl-CoA, restoring fatty acid synthesis and male feritilty (Ruiz and Daniell, 2005). 4.2.2. p-hydroxybenzoic acid Bacterial and synthetic genes have been explored for polymer production applica- tions in plastid transformants. The shikimate pathway, responsible for the synthesis of aromatic amino acids (reviewed in Herrmann and Weaver, 1999), has gained considerable attention from biotechnology through the years. Inhibition at chloro- plast localized enolpyruvylshikimate-3-phosphate synthase (EPSPS) is the activity that makes Roundup® constituent glyphosate (N-phosphonomethylglycine) an effective herbicide. Resistance to glyphosate is engineered into a plant by adding a gene from a soil bacterium (ie Agrobacterium sp), or even another plant species, that encodes a version of the EPSPS that is resistant to the herbicide (reviewed in Dill, 2005). The last of seven steps in the shikimate pathway (which occurs exclusively in plastids and bacteria) results in chorismate. This compound is of particular interest as it provides abundant substrate for a bacterial enzyme which has been expressed in transplastomic plants to produce p-hydroxybenzoic acid (pHBA). The E. coli ubiC gene encodes chrosimate pyruvate-lyase (CPL) which catalyzes the conversion of chorismate to pHBA, the major monomer in liquid crystal polymers. Introducing this gene to the plastid genome resulted in accumulation of pHBA polymer to extraordinary levels, up to 26.5% of the dry weight (Viitanen et al., 2004). Although this represents a 50-fold enhancement of the best values reported for nuclear trans- formants transplastomic plants were indistinguishable from wild types. 4.2.3. Protein based polymers Naturally occurring proteins, such as the silks of insects and mammalian elastin, exhibit properties which make their controlled production desirable. Genetically engineered expression of protein based polymers (PBPs) offers the potential to precisely program attributes like molecular mass and amino acid composition that dictate the functional properties of the end product (Meyer and Chilkoti, 2002). In order for the production of PBPs for industrial applications to be feasible in transgenic plants, high levels of accumulation will have to be obtained. The synthetic variations of PBPs designed and tested to date have demonstrated an extraordinary degree of biocompatibility making them attractive for medical applications (Betre et al., 2002). The protein sequence of mammalian elastin, one of the strongest known natural fibers, consists of repeated blocks of five amino acids. It is from this sequence that synthetic analogs are designed. One such analog is the protein based elastomer GVGVP121 (amino acids Gly-Val-Gly-Val-Pro repeated 121 times). This concatomer sequence has been expressed in several systems including E.coli (Daniell et al., 1997) Aspergillus nidulans (Herzog et al., 1997) tobacco NT1  PLASTID PATHWAYS 99 cells (Zhang et al., 1995), and whole tobacco plants (Zhang et al., 1996; Guda et al., 2000). One attractive feature of these types of polymers is that they display inverse temperature transition (Urry, 1988). While GVGVP121 is soluble in water at room temperature it forms a more ordered aggregate as temperature increases. This reversible process could lend itself to the eventual purification of the polymer from plant tissues. Although chloroplast transformants expressing GVGVP121 accumu- lated up to 100-fold more transcript for the polymer than the nuclear transformant, the polymer itself was not abundant. Western analyses revealed polypeptides of a smaller molecular weight than GVGVP121. Because this sequence contains no known protease cleavage sites the authors reasonably suggest that these products were the result of incomplete polymerization. Certainly transcription was not limiting. The unusually heavy demand for constituent amino acids may have been the factor hindering continued polymer synthesis through all 121 repeats. Glycine, which comprises 40% of the final sequence regardless of length, is synthesized in mitochondria by two mechanisms. Serine may be converted to glycine by the action of Serine Hydroxymethyl Transferase in what amounts to a reversal of serine synthesis. Thought to predominate is the synthesis of glycine from CO2 and NH4 with N5 N10 -mthylene tetrahydrofolate (THF) acting as a donor of one carbon units via glycine synthase. Could these activities be established in plastids coupled to polymer expression? The ability to engineer such constructs for incor- poration in the plastome in a single transformation event makes it a feasible investigation. 5. PROSPECTS FOR ENHANCING PLANT PRODUCTIVITY Chloroplast biotechnology has allowed the examination of a variety of plastid functions that prior to its inception had seemed elusive. Ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco) has attracted considerable attention and presents a target for metabolic engineers interested in improving agricultural productivity (Spreitzer and Salvucci, 2002). One way to increase the radiation use efficiency (RUE) in plants could be to enhance Rubisco’s affinity for CO2 over its competitor O2 in reaction with ribulose-1,5-bisphosphate (RuBP). A 20% increase in light saturated net carbon exchange (Amax ) could result from doubling the enzymes affinity for CO2 in the reaction and abating photorespiration (Reynolds et al., 2000). Neither substrate binds directly to the enzyme making manipulation of Rubisco for speci- ficity especially challenging (reviewed in Parry et al., 2003). Further complicating genetic modification of the hexadecameric holoenzyme has been the location of the genes for the large (rbcL) and small (rbcS family) subunits in the plastid and nucleus respectively (Rodermel, 1999). The large subunit (LSU), where the active site is located, is well conserved in crop plants (∼90% identity) whereas the small subunits (SSU), to which a function has not yet been ascribed, are less so (∼ 70% identity). The divergence seen in the small subunit may account for variation in efficiency among Rubiscos from different species (Spreitzer, 1999). Several approaches have been undertaken to introduce subunit genes from cyanobacteria, algae, sunflower and 100 RUHLMAN AND DANIELL pea to tobacco or Arabidopsis through nuclear and plastid transformation (reviewed in Parry et al., 2003). A recent effort to recover Rubisco activity in tobacco plants with antisense-ablated SSU expression (Rodermel et al., 1988) has demonstrated that, with the use of appropriate regulatory signals, it is possible to effectively relocate rbcS to the plastid genome in a land plant. Employing rbcS constructs fused to T7g10 RBS or the psbA 5’ UTR facilitated SSU protein accumulation to 60% and 106% of the wild type control respectively. Moreover these transplantomic plants not only accumulated abundant SSU, they were able to fully complement the antisense phenotype assembling functional Rusbisco (Dhingra et al., 2004). 5.1. Future Perspectives 5.1.1. Sequencing crop plastomes The future of chloroplast biotechnology for a range of applications, including metabolic engineering gives us reason to be optimistic. There are however several areas of concern that will require attention if the potential of this technology is to be realized. Where it was once thought that plastome sequences held little variation from one species to the next, recently sequenced genomes are revealing a richness of diversity among plastid genomes that was not expected (Kanamoto et al., 2005; Kim et al., 2005; Daniell et al., 2006). While overall gene content and order is highly conserved in land plants this same conservation is not observed in non-coding sequence such as introns and intergenic spacers (IGS) which, along with the UTRs of genes comprise up to 50% of the plastome (Saski et al., 2005; Daniell et al., 2006; Jansen et al., 2006; Lee et al., 2006). Foreign genes are targeted to IGS regions as not to disrupt endogenous gene function. Integration of foreign sequences is dependent on homologous recombination between the transformation vector and the plastid genome. It is possible to achieve integration without complete homology but recombination and hence transformation efficiency is impaired (DeGray et al., 2001; Zubko et al., 2004). Additionally, evalu- ations of UTRs from different species indicates the need to employ species specific regulatory elements such as promoters and translation sequences to elevate the level of foreign protein expression (Kramzar et al., 2006). The last year has seen a remarkable effort in this regard. Grevich and Daniell (2005) reported just six crop genomes had been published. Since then a number of agronomically important genomes such as cotton (Lee et al., 2006; Ibrahim et al., 2006), coffee (Samson et al., 2007) grape (Jansen et al., 2006), carrot (Ruhlman et al., 2006), cucumber (Kim et al., 2006), sweet orange (Bausher et al., 2006), potato (Chung et al., 2006) and tomato (Daniell et al., 2006), barley, sorghum and turfgrass (Saski et al., 2007) have been added (see a complete list in http://megasun.bch.umontreal.ca/ogmp/projects/other/cp.list.html) and methods have been published to enable expansion of this list (Dhingra and Folta, 2005; Jansen et al., 2005). 5.1.2. Alternative methods of DNA delivery If this interest in plastid genomics stems from a desire to bring new crop species to chloroplast biotechnology it will need to be accompanied by an understanding  PLASTID PATHWAYS 101 of the somatic embryogenesis system to enable selective pressure sufficient to establish homoplasmy in systems where adventitious shoot formation is not an option for regeneration (Daniell et al., 2005a). The elucidation of protoplast to plant culture techniques may facilitate plastid transformation in recalcitrant species. The use of polyethylene glycol to introduce foreign DNA into plant chloroplasts has proven feasible in isolated protoplasts of lettuce and tobacco (Eibl et al., 1999; Lelivelt et al., 2005). 5.1.3. Eliminating antibiotic resistance markers Ultimately acceptance of regulatory agencies and the marketplace will demand the elimination of antibiotic resistance markers in transgenic plants. Several approaches are being explored in this regard. Implementation of the CRE recombinase system in plastid transformants takes advantage of functional lox sites within the plastome (Corneille et al., 2001). While this system was highly efficient in marker elimination it relies on the use of nuclear transgenic plants harboring the recombinase gene as opposed to the wild type background. Another system which relies on recombination for marker removal employs engineered direct repeats flanking the sequence to be discarded. Marker excision by this method was automatic and highly efficient without the need to engineer two genomes (Day et al., 2005). Ideally an operative system that utilizes native plant sequences for markers, such as betaine aldehyde dehydrogenase (BADH) from spinach (Daniell et al., 2001c), rather than antibiotic resistance would be employed. Work should be encouraged to identify possible positive selective strategies incorporating environmentally benign markers. Transformation of crop plant plastids holds promise as a means to address some of the many and varied needs of the global population. What could be accomplished through the insightful application of this technology may only be limited by our own preconceptions of what is possible. Among the many benefits associated with plastid transformation, the ability to cointegrate several genes as an operon is especially attractive to metabolic engineering. What is the maximum number of genes, or length of sequence that can be transferred to and subsequently expressed in transgenic plastids? Presently we do not know what the upper limit. When considering the possibilities keep in mind the words of William Blake: “What is now proved was once only imagined…” REFERENCES Adam Z, Clarke AK (2002) Cutting edge of chloroplast proteolysis. Trends Plant Sci 7: 451. Alergand T, Peled-Zehavi H, Katz Y, Danon A (2006) The chloroplast protein disulfide isomerase RB60 reacts with a regulatory disulfide of the RNA-binding protein RB47. Plant Cell Physiol 47 (4): 540–548. Amira G, Ifat M, Tal A, Hana B, Shmuel G, Rachel A (2005) Soluble methionine enhances accumulation of a 15 kDa zein, a methionine-rich storage protein, in transgenic alfalfa but not in transgenic tobacco plants. J Exp Bot 56: 2443–2452. Arlen PA, Falconer R, Cherukumilli S et al. (2007) Field production and functional evaluation of chloroplast-derived interferon alpha 2b. Plant Biotechnol J doi:10.1111/j.1467–7652.2007.00258.X