Development of a novel rep-inducible tomato leaf curl virus expression system

Pathogen-derived resistance (PDR) strategies, particularly those based on post-transcriptional gene silencing, have been used with great success for the generation of transgenic plants with resistance to RNA viruses. In contrast, a suitable strategy for transgenic resistance to ssDNA plant viruses, including those viruses belonging to the Geminiviridae, has remained elusive. Further, there is no convincing evidence that either post-transcriptional gene silencing, or pathogen-derived resistance in general, would be broadly applicable to ssDNA plant viruses. Researchers at QUT have been developing a novel resistance strategy against ssDNA viruses based on virus-activated expression of a stably integrated suicide gene. The strategy, based on InPAct (In Plant Activation) technology, relies on a &quotsplit" suicide gene cassette being arranged in such a way that expression of a lethal ribonuclease (barnase) is dependent on the virus-encoded replication-associated protein (Rep). Upon infection, Rep mediates the release of the construct resulting in the reconstitution of a transcribable and translatable episomal suicide gene expression cassette. The research for this PhD describes the development of an InPAct vector designed to confer resistance to Tomato leaf curl begomovirus (ToLCV), a major cause of disease in Solanaceous crops in the tropics and subtropics.\ud \ud \ud \ud ToLCV-based InPAct vectors were constructed based upon two ToLCV isolates from Australia and North Vietnam. Prior to the generation of InPAct cassettes, the entire ToLCV-[Au] and ToLCV-Vie intergenic regions (IRs) were embedded within the castorbean catalase intron of a β-glucuronidase expression vector to determine the effect of the IR upon transcript processing. Using transient reporter gene assays in tobacco NT-1 cells, it was demonstrated that the ToLCV IRs both contained cryptic intron splice sites which interfered with efficient transcript processing and GUS expression. A series of truncations to the IRs were subsequently made to identify the potential cryptic intron splice sites and/or interfering sequences in both the ToLCV-[Au] and ToLCV-Vie IRs. The final truncated IRs, which were used in the construction the InPAct cassettes, comprised approximately 100 bp and appeared to contain all the necessary cis-acting elements required for efficient rolling circle replication (RCR). Using histochemical GUS assays and Southern analyses, the InPAct cassettes were shown to be activated and replicated only in the presence of the cognate viral Rep. GUS expression levels were shown to be further enhanced in the presence of the ToLCV replication-enhancer protein (REn) and by the addition of the Tobacco yellow dwarf mastrevirus origin of second strand synthesis into the cassette. Under these conditions, Rep-activated GUS expression from the InPAct vectors was found to reach levels similar to that of the benchmark CaMV 35S promoter. \ud \ud \ud \ud Fifteen independent transgenic lines containing the ToLCV-[Au] and -Vie InPAct-GUS cassettes were generated by Agrobacterium-mediated transformation of tobacco leaf discs. Using agroinfiltration and histochemical assays, Rep-mediated activation of the InPAct cassettes and subsequent GUS expression was demonstrated in 11 out of the 15 lines tested; six of which showed expression levels equivalent to, or higher than, that obtained using a CaMV 35S promoter control. Evidence for activation of the integrated InPAct cassettes at the molecular level was provided by Southern analyses, with showed both linear and open circular forms of the replicating InPAct episome in genomic DNA extracted from infiltrated leaf tissue. \ud \ud \ud \ud Following the demonstration of Rep-activatable reporter gene expression and episomal replication of the ToLCV-based InPAct-GUS vectors using transient and stable tobacco transformation assays, new ToLCV-based InPAct vectors were designed to express the lethal RNase, barnase, in an attempt to generate virus resistant plants. Although transient assays in NT-1 cells demonstrated some &quotleaky" expression of barnase from the InPAct vectors, the level of barnase-mediated cell death from the InPAct vectors was found to be significantly increased in the presence of the cognate Rep and REn. Thirteen independently transformed tobacco lines containing the ToLCV-[Au] InPAct-barnase cassette were generated by Agrobacterium-mediated transformation of tobacco leaf discs. However, agroinfiltration of these plants with ToLCV Rep and REn failed to activate a barnase response. Subsequent molecular analyses on two transgenic lines revealed that both contained mutations in the barnase-coding gene in a region known to encode the active site. These mutations were presumed to result from the leaky barnase expression during initial stages of the Agrobacterium transformation which would favour the selection of barnase mutant InPAct plants. \ud \ud \ud \ud To overcome the problems associated with leaky expression of barnase, a barstar-expression cassette was included in the ToLCV-[Au] InPAct-barnase cassette. Transient assays in non-transgenic tobacco leaves demonstrated that the basal levels of barstar expressed from the modified InPAct vector were sufficient to negate the effects of leaky barnase expression. Importantly, however, the level of barnase expression in the presence of Rep and REn was shown to be sufficient to overcome the basal levels of barstar. Seventeen independently transformed lines were generated with the ToLCV-[Au] InPAct-barnase/barstar cassette, and analysis of one line revealed the presence of an uncorrupted barnase-coding region. Using transient agroinfiltration assays, seven of the transgenic lines showed varying levels of cognate Rep and REn-activated, barnase-induced cell death.\ud \ud \ud \ud Fifteen transgenic lines were challenged with ToLCV-[Au] by injection of recombinant Agrobacteria containing an infectious ToLCV clone. Unfortunately, all lines displayed typical ToLCV symptoms and tested positive for virus by PCR at 28 days post-inoculation. The inability of the InPAct cassette to confer resistance to ToLCV may have been due to one or a combination of factors, including (i) a delay in barnase-induced cell death, (ii) homology-dependent silencing of the integrated cassette, (iii) generally low-level, Rep-activated barnase expression or (iv) excessive virus load due to the artifical method of inoculation.\ud \ud \ud \ud This study details the first report of a ToLCV-based InPAct system for Rep-induced transgene expression in planta. Despite failing to generate ToLCV-resistant plants, the research findings will provide a solid foundation to develop a more effective InPAct vector and ultimately assist in the generation of transgenic plants with resistance to ToLCV and potentially other ssDNA plant viruses, particularly the begomoviruses