Application of hordothionins and cecropin B for engineering bacterial disease resistance into plants

Bacterial diseases can cause a drastic decrease of yield in certain crops. Breeding for bacterial disease resistance therefore is of utmost necessity. Up to now, traditional plant breeding was the only method to reach this goal. Recent developments in genetic engineering technology however provide novel ways for the production of disease resistant plants. This thesis describes the results of two research projects that have been undertaken to investigate the potential of such a novel way, namely the introduction and expression of genes coding for antibacterial proteins in plants. In the first project, the potential of the hordothionins from barley (Hordeum vulgare) endosperm, has been investigated, and in the second, the potential of cecropin B from the giant silkmoth (Hyalophora cecropia).In the first part of chapter 1, the literature available on plant thionins is presented. General information on the different thionin types, homology, occurrence in nature, molecular structure and toxicity for microorganisms and cultured cells is listed. Since their role in nature has not yet been established, although a putative role in plant defense has been proposed by several groups, special emphasis is put on the numerous divergent activities displayed by the different thionins in conjunction with possible modes of action and biological roles. Data collected from literature indicate that thionins might expose their toxic activity in vitro by several mechanisms: by acting as a thiol intermediate in reducing and oxidizing proteins or by direct binding to DNA and/or RNA or by interaction with the phospholipid membrane, acting most likely on Ca 2+-channels or -pumps, and/or Ca 2+-ATPases. In the second part of chapter 1, an overview of the literature on insect cecropins is presented. General information on molecular structure, toxicity and known mode of action of these proteins is presented. Special emphasis is put on one of them, cecropin B, which was under investigation in the second research project.Our first choice was to investigate the feasibility of using thionin encoding sequences for engineering bacterial disease resistance into solanaceous crops. The hordothionins originating from barley endosperm were chosen because of the availability of nucleotide and amino acid sequences. To establish the potential of hordothionins, the toxicity for plant pathogenic bacteria had to be determined first. To this end the thionins from wheat and barley endosperm were isolated (chapter 2). The thionins were purified in few steps from flour by petroleum-ether extraction and hydrochloric acid treatment of the resulting lipoprotein, followed by ion-exchange chromatography. The hordothionins were separated into two forms, HTH-1 and HTH-2, probably reflecting the two forms described in literature, namely α- and β-hordothionin. In vitro experiments indicated that purothionin, hordothionin and HTH-1 and HTH-2 were equally toxic to Clavibacter michiganensis subsp. michiganen sis, the causal agent of bacterial canker on tomato, C.michiganensis subsp. sepedonicus, the causal agent of ring rot on potato and Xanthomonas campestris pv. vesicatoria, the causal agent of a foliage and fruit spot disease on tomato and pepper. Erwinia spp. and Pseudomonas spp. were insensitive.Since hordothionins appeared to be toxic for a number of bacteria that affect tomato and potato, several hordothionin encoding constructs were made (chapter 3). Analysis of published hordothionin cDNA clones indicated that the 5 kD mature hordothionin was made as a much larger precursor protein. This precursor consisted of an amino-terminal signalpeptide, followed by the mature hordothionin exhibiting antibacterial activity, and a carboxy-terminal so-called acidic peptide of unknown function. Since no such cDNA or genomic hordothionin clones were available to us, we decided to chemically synthesize the genes making use of the many advantages of synthetic genes. In chapter 3, the design and construction of seven different α- and β-hordothionin encoding constructs is described. Gene constructs were made to study expression in cytosol and secretion into the apoplast. Genes were designed for optimal expression in solanaceous crops by adapting the codon usage and optimizing the translation initiation region, and for convenience in subsequent cloning steps.The seven hordothionin gene constructs made were cloned in a plant expression vector under the control of the constitutive cauliflower mosaic virus 35S promoter and introduced in tobacco to study their expression, processing of precursors, sorting and biological activity (chapter 4). Analysis of a large number of transgenic plants indicated that the signalpeptide was essential for expression, whereas the acidic peptide facilitated transport of mature hordothionin and increased accumulation at least tenfold compared to plants harboring constructs without this acidic peptide coding sequence. Fractionation of protoplasts prepared from transgenic plants indicated that hordothionin accumulated in the microsomes and membranes and was not secreted into the medium. In addition, hordothionin was not secreted into the apoplast in intact leaves. The hordothionin partially purified from leaves of transgenic tobacco plants exhibited in vitro toxicity for C.michiganensis subsp. michiganen sis, at comparable doses as the hordothionin from barley endosperm. The tobacco phytopathogen P.syringae pv. tabaci was however not affected in growth in transgenic tobacco plants exhibiting high hordothionin expression levels (chapter 4).In vitro growth inhibition experiments indicated clearly that hordothionin was toxic for a few bacteria that were pathogenic on tomato and potato. For this reason, the best performing construct was introduced in tomato (chapter 5). Transgenic plants were selected exhibiting high hordothionin expression levels, selfed to obtain plants homozygous for the transgene and evaluated for resistance. No differences were observed between control plants and transgenic tomato plants exhibiting high hordothionin expression levels upon infection with C.michiganensis subsp. michganensis. In addition, no growth inhibition of X.campestris pv. vesicatoria was observed in leaves of these transgenic plants upon infiltration of a bacterial suspension. However, less symptoms were visible on these transgenic plants upon spray-inoculation with a suspension of the latter.Since hordothionins were not toxic to Erwinia spp. and Pseudomonas spp. in vitro (chapter 2), a second antibacterial protein was chosen exhibiting toxicity to these bacteria. Three different cecropin B gene constructs were made and introduced in tobacco. Analysis of transgenic plants indicated that the genes were transcribed into mRNA, however the protein could not be detected (chapter 6). By mixing a synthetic cecropin B peptide with different tobacco cell extracts, it appeared that the peptide was rapidly degraded, whereas boiling of the extracts prior to mixing did not result in cecropin B degradation, suggesting protease degradation. This was confirmed by experiments which indicated that the protease inhibitors chymostastin and to a lesser extent PMSF, also inhibited degradation. Transgenic plants exhibiting high cecropin B-mRNA levels were nevertheless evaluated for resistance to two pathogens, P. solanacearum and P. syringae pv. tabaci. Results from these experiments clearly indicated that transgenic plants were not resistant.Finally, in chapter 7 a general discussion on the topic described in this thesis is presented. Also, other potential approaches to obtain bacterial disease resistant plants are discussed and other applications of hordothionin encoding sequences.