TOWARDS PLASTID TRANSFORMATION IN MAIZE
Ralph Bock
December, 2007

Transgene expression from the plastid (chloroplast) genome offers unique attractions to plant biotechnologists, including high-level foreign protein expression (presumably due to the high copy number of the plastid genome per cell), absence of epigenetic effects, and the convenient possibility of expressing multiple transgenes by linking them in operons¹. In addition, transplastomic technology greatly improves gene containment due to the maternal inheritance of plastids and their genomes in most crop species, which drastically reduces transgene transmission through pollen. However, broad application of plastid genome (plastome) engineering in agricultural biotechnology has been hampered by the lack of plastid transformation protocols for major crops. Unfortunately, cereals, the world's major food crops, are a particularly recalcitrant species; thus, in spite of enormous efforts, progress has been very limited^2,3 and there are still no protocols available for the production of stable transplastomic plants in any cereal species.

The availability of efficient methods for plant tissue culture and regeneration as well as the development of stringent selection protocols for transplastomic lines represent the main bottlenecks in plastid transformation. Leaf material is the preferred source material for plastid transformation, because it is rapidly produced in large amounts and allows multiple successive rounds of selection and regeneration. The latter is of utmost importance to the development of workable plastid transformation protocols, because primary plastid transformants are heteroplasmic, that is, contain large amounts of residual wild type plastid genome copies. To obtain genetically stable transplastomic lines (i.e., plants with transgenic plastid genomes), primary transformants must be purified to a state of homoplasmy by repeated cycles of plant regeneration under selective pressure^4,5. The impossibility of conducting such successive regeneration rounds in monocot species due to the lack (or extremely low efficiency) of leaf-based regeneration systems is the main reason for the failure to generate genetically stable chloroplast transformants in cereals. While rice chloroplast transformants could be readily obtained, all lines remained heteroplasmic and eventually lost the plastid transgenes in the absence of a protocol for carrying out repeated regeneration cycles².

A leaf-based callus induction and plant regeneration system for maize
We sought to overcome the limitations of the existing tissue culture systems by developing a plastid transformation protocol for maize (Zea mays L.). Currently, the source material used in maize (nuclear) transformation experiments comes nearly exclusively from callus cultures induced from immature zygotic embryos and propagated in the dark. We were interested in establishing a tissue culture and plant regeneration system that uses maize leaf tissue and thus is independent of zygotic embryos and greenhouse facilities. Earlier work had shown that cereals can undergo callus induction and plant regeneration via somatic embryogenesis from leaf explants, although the efficiency is often very low.

To develop an efficient leaf-based system for maize, we first systematically tested leaf explants of differing size, age, and developmental stage for their responsiveness to callus induction on different tissue culture media. An approximately 2 cm long section at the leaf base of young seedlings (Fig. 1A) sectioned into 1-2 mm × 1 mm pieces (Fig. 1B) was found to provide the most suitable source material for callus induction from maize leaves. In addition to optimizing the concentrations for known critical components of maize tissue culture media, such as proline, the auxin 2,4-dichlorophenoxyacetic acid (2,4-D), and silver nitrate, we tested a number of other chemicals for their effect on callus induction from leaves and found the polyamine spermidine has a stimulatory effect. The fully optimized protocol resulted in efficient callus induction from leaf segments, with a substantial proportion of yellow embryogenic type I-like callus⁶ (Fig. 1C,D). This callus type proved to be highly regenerable (Fig. 1E,F) and produced plants that grew normally and were fully fertile (Fig. 1G,H).

Figure 1. Procedures for callus induction from young maize leaves and plant regeneration. (A) Harvest of leaf tissue from young seedlings (Pa91×H99 hybrid) grown under aseptic conditions. The shoot explant taken to set up callus cultures from leaf pieces is indicated by the brace. (B) Excised leaf pieces exposed to callus-inducing medium. (C) Callus induction after incubation in the dark for five weeks. A consecutive series of leaf pieces from the leaf base (upper right corner) to the tip (lower left corner) is shown. (D) A type I-like yellow callus directly induced from a leaf explant (visible as brown tissue at the bottom). (E) Shoot induction from leaf-derived calli after incubation on a regeneration medium containing 0.5 mg/l benzylaminopurine (BAP) for four weeks in the light. (F) Simultaneous shoot and root initiation from leaf-derived calli by incubation for three weeks in the light on a regeneration medium containing 2 mg/l naphthyl acetic acid (NAA). (G) Growth of a regenerated maize plant to maturity in the soil. (H) Seed production of regenerated maize plants. Regenerated plants that had passed through the regeneration procedure were fertile and seed production was indistinguishable from plants that were obtained directly from seeds. Modified from Transgenic Res. 16, 437-448 (2007).

The optimized culture protocol was also tested on tissues other than leaf bases. In these experiments, immature tassels harvested from greenhouse-grown plants performed best, because they produced embryogenic callus at very high frequency, which was exclusively type I-like and highly regenerable.

Nuclear transformation of leaf-derived calli
To test whether the type I-like calli induced from leaf pieces provide suitable material for stable maize transformation, genetic transformation experiments were performed using the biolistic protocol. A metabolic marker was employed for the selection of transgenic lines: the phosphomannose isomerase gene (manA or pmi), whose expression allows callus growth on mannose. As mannose cannot be metabolized by maize cells, non-transformed cells die from starvation, whereas transgenic cells expressing the pmi marker gene can grow by utilizing mannose as the sole carbon source.

Particle gun-mediated transformation with a plasmid carrying a pmi cassette followed by selection of transformed cell lines on mannose-containing medium yielded one to four transgenic clones per bombardment. Molecular analyses confirmed that cell lines growing on mannose were indeed transformed. Transgenic calli could be readily regenerated into plantlets, and most of them produced phenotypically normal and fertile maize plants⁶, confirming that leaf-derived callus provides excellent material for transformation experiments.

Induction of regenerable callus from leaves in the light
In maize plastid transformation experiments, it is highly desirable to perform selection and regeneration of transplastomic clones in the light. This is because the aminoglycoside antibiotics used for selection of chloroplast transformants in dicots are nearly ineffective on maize callus growing in the dark. Unfortunately, the most commonly used antibiotic, spectinomycin, cannot be used at all, because maize cells are endogenously resistant to this drug. Streptomycin or kanamycin, two other antibiotics that have been demonstrated to facilitate the selection of chloroplast transformants in dicots, do not significantly inhibit the growth of maize callus in the dark⁶. However, when calli were exposed to regeneration medium containing either streptomycin (100 mg/l) or kanamycin (50 mg/l) in the light, growth and regeneration were strongly inhibited, suggesting that effective selection for chloroplast transformants with streptomycin or kanamycin necessitates incubation in the light.

We therefore sought to develop a leaf-based tissue culture system that would permit selection of transgenic cell lines in the light, which complicates monocot tissue culture, for example, by precluding the inclusion of silver nitrate in the medium. By testing a series of compounds for their possible stimulatory effects on callus initiation from maize leaf explants in the light, we identified the phytohormone phytosulfokine-alpha (α-PSK) as a substance that strongly promoted embryogenesis from leaves in the light. α-PSK is a sulfated pentapeptide, which originally was discovered as a regeneration-stimulating phytohormone in rice. Addition of very small amounts of α-PSK (30 to 75 nM) was sufficient to elicit the stimulatory effect on callus induction from maize leaf pieces in the light.

Prospects for plastid transformation in corn
The commercialization of transgenic corn varieties expressing herbicide tolerance genes and/or genes conferring insect resistance has raised serious environmental concerns, many of which are related to the risk of unwanted transgenes spreading via pollen. These concerns affect the co-existence of genetically engineered and conventional crops as well as the risk of inadvertently contaminating the food chain with transgenes for biopharmaceuticals or industrial products. As maize is one of the world's most important staple foods and is considered a promising production platform for molecular farming, the development of a plastid transformation technique for maize is particularly desirable.

In principle, two sources of material can be used in plastid transformation experiments: callus cultures or leaf tissue. The use of conventional maize callus cultures seems impractical for two reasons. First, as explained above, selection for the antibiotic resistances established for plastid transformation in dicots (spectinomycin, streptomycin, and kanamycin) either does not work at all in corn or is extremely inefficient when done in the dark. Second, successful plastid transformation requires multiple rounds of regeneration and selection to eliminate wild type plastid genomes and achieve a homoplasmic transplastomic state⁴, an essential prerequisite for obtaining genetically stable plastid-transformed plants. These multiple cycles of selection and regeneration are usually conducted by taking leaf explants from primary transformants and subjecting them to a new regeneration round under stringent antibiotic selection^4,5. In fact, the impossibility of carrying out these additional regeneration rounds in cereals is considered the major limitation to the implementation of chloroplast transformation in maize and rice². While chloroplast-transformed rice cell lines could be obtained at reasonable frequencies, all attempts to stabilize the lines and purify them to homoplasmy have failed^2,3.

Thus, the advantage of the new leaf-based system⁶ for the development of plastid transformation in corn is twofold: (i) Using the α-PSK protocol, selection for chloroplast transformants can be done in the light, making antibiotic selection much more effective; and (ii) primary chloroplast-transformed plants can be subjected to additional rounds of regeneration, which should facilitate the isolation of genetically stable homoplasmic plant lines. In fact, plants regenerated from leaf-derived callus could be easily regenerated again from leaves into fertile plants, demonstrating that the system is suitable for performing multiple rounds of selection and regeneration.

This progress not withstanding, there is probably still some way to go until plastid transformation in maize (and other cereal crops) will become a reality. The construction of suitable maize-specific plastid transformation vectors and the optimization of the in vitro selection procedures (using streptomycin and/or kanamycin as selecting antibiotics) represent the next challenges.

Ralph Bock
Max-Planck-Institut für Molekulare Pflanzenphysiologie
Am Mühlenberg 1, D-14476 Potsdam-Golm, Germany
rbock@mpimp-golm.mpg.de; fax: 49-331-567-8701