Correlation of plastid gene expression with albinism

In a large number of plant species, the in vitro regeneration of doubled haploid plants from immature pollen enables the production of completely homozygous plants within a single generation. This method has the potential to speed up plant breeding programmes and is also an important tool for basic research on pollen development and embryogenesis. In cereals, however, microspore embryogenesis is frequently accompanied by the generation of pigment-deficient (albino) plants, which are unable to perform photosynthesis. From previous studies this albino phenotype was thought to be the result of mutations and rearrangements in the plastid genome. In the preceding project we could show that plastid DNA rearrangements are not the primary reason for the generation of an albino phenotype. Instead, we demonstrated that in all cases albino plants show a deviant transcript pattern of plastid- encoded genes and a general deficiency in plastid translation irrespective of the presence or absence of rearrangements in the plastid genome. The molecular reason for the albino phenotype therefore has to be found in the control of transcriptionranslation and the mechanism of generating plastid genome rearrangements might be related to such translation deficiencies. As a consequence of these results we propose to address the following questions in the current project: At which stage is the plastid development blocked in albino plants, what is the reason for a general translation deficiency in plastids of albino plants, do specific features in the cereal plastid genome lead to an instability of plastid genomes during in vitro conditions, what is the reason for recombinational events leading to plastid DNA deletions in albino plants, is there a connection between the general translation deficiency observed in all albino plants and the occurrence of plastid DNA deletions, is translation deficiency already found in developing pollen, what is the role of the nuclear genome in albino plant formation (translation deficiency, unstable plastid genome). Answers to these questions will help to improve doubled haploid plant production methods in cereals. Furthermore, they can function as a tool for a better understanding of fundamental mechanisms in plastids, like chloroplast development, signalling between plastid and nucleus, the role of ORFs retained in the plastid genome and regulation of plastid gene expression.

It has been reported that the hypothetical tobacco chloroplast gene ycf1 has similarity to a bacterial gene, which is necessary in DNA repair, replication and recombination. This indicates that ycf1 might work as a stabilising factor in chloroplasts especially under stress conditions. To address the question whether the tobacco ycf1-gene or the truncated, non protein-encoding, ycf1-gene detected in cereals (ycf1 pseudogene) correlates with the formation of albino plants in cereals during doubled haploid production (stress induced production), the ycf1-locus in tobacco was modified. Using chloroplast transformation vectors, three different alleles were inserted. First, the entire ycf1- gene was knocked out with a marker gene to obtain information about the function of the ycf1-gene (KO-plants). Alternatively, the tobacco ycf1-gene was replaced with the ycf1 pseudogene of rice in such a way that either the pseudogene could be transcribed (RP-plants) or that transcription of the pseudogene was inhibited (TI-plants). These replacement experiments should provide information whether the pseudogene destabilises the chloroplast genome during stress induced doubled haploid production. All three alleles regenerated positive transgenic tobacco plants. Although all transgenic plants contained both, the wild type ycf1-allele and the ycf1-pseudogene, the RP-plants displayed a phenotype. Plants containing the other two alleles, KO or TI, only displayed a wild type phenotype. RP-plants showed yellowish-white/bright-green leaves with green spots. They grew very slow and reached only half the size of a normal tobacco plant. White leave-sections displayed lesions that led to degradation of the affected tissue. Since transcription of the ycf1 pseudogene is only possible in the RP-plants the phenotype might result due to an interaction of transcripts from the wild type ycf1 and the inserted pseudo-ycf1. On the other hand, a transfer of a protein-encoding ycf1 from the chloroplast to the nucleus cannot be excluded, because gene transfer between different genomes is a still ongoing, evolutionary process. However, molecular studies of nuclear genomes of rice and tobacco did not detect a protein-encoding ycf1-gene, similar to the tobacco ycf1-gene in the rice nuclear genome. In addition, several copies of the rice pseudo-gene were detected in the rice nuclear genome. Summarising, tobacco has a functional ycf1-gene in the chloroplast genome and no additional pseudogenes whereas rice does not have a functional ycf1-gene neither in the chloroplast nor in the nuclear genome but therefore several copies of pseudogenes in the nucleus.