STUDI FINALIZZATI AD AUMENTARE L’ESPRESSIONE DI TRANSGENI IN ENDOSPERMA DI RISO MEDIANTE INGEGNERIZZAZIONE DI PROMOTORI CHIMERICI SEME-SPECIFICI

The biopharmaceutical industry is seeking novel, more performing platforms for the production of recombinant proteins. In the last decade plants have been emerging as a good alternative to animal cells in the manufacturing of pharmaceutical proteins, especially for injectable preparations. Several studies have shown that the seed-specific expression of a transgene is more convenient than the constitutive expression. Specifically, the endosperm of cereals can be a suitable tissue to accumulate and durably store high amounts of heterologous proteins. Oryza sativa is a promising species for biotechnological application, but up to this time, the yield of recombinant protein is still not enough to push industries toward this innovative manufacturing platform. One of the more effective ways to improve the yield of a recombinant protein is to deal with the promoter. The rice endosperm-specific expression can be easily achieved by the use of a variety of promoters of the rice seed storage proteins (RSSPs). These promoters have been widely studied and some of the cis- and trans-acting factors responsible of the specific expression have been identified. Several authors have described methods for promoters sequence analysis and manipulation. However, no studies had been published about the engineering of promoters for rice endosperm expression. The aim of this PhD has been the application of those manipulation methods to the RSSP promoters. An early goal has been the definition of methods for sequence analysis and manipulation in order to increase the expression level of the RSSP promoters. The ultimate purpose of this PhD has been the construction of a synthetic promoter with activity higher than the natural homologues. The research project had been structured according to an iterative process and two experimental cycles had been planned. Fourteen sequences of natural RSRP promoters have been analysed, with the purpose of identifying the cis-elements. Six of these promoters have been selected on the base of sequence analysis and previously published data. The selected promoters drive the expression of the following RSSPs: B1, B4 and C glutelins, 10 kDa and 16 kDa prolamins and 26 kDa globulin (with acronym: GluB1, GluB4, GluC, Prol10, Prol16 and Glb, respectively). These promoters have been cloned following PCR isolation from Oryza sativa (var. CR W3) genome. A number of chimera sequences have been designed by combining in various ways portions of these promoters. Seven of these chimera sequences have been selected for in vivo testing. In order to estimate the in vivo performances, the synthetic promoters have been inserted in an expression cassette with a reporter gene coding for a phytotoxic protein, namely the human acid β-glucosidase (GCase). The only variable of the experiment has been the sequence upstream the TATA-box. The synthetic promoters have been tested in transformed plants of Oryza sativa (var. CR W3) together with three of the strongest natural RSSP promoters, i.e. GluB4, Glb and GluC. A population of at least 20 individuals has been produced for each construct. The amount of the reporter protein in the seed of each plant has been determined and each promoter has been evaluated on the base of three parameters: maximum expression level, average expression level calculated taking account the whole population or the first quartile. Among the three natural promoters, GluB4 has proven to be the strongest natural RSSP promoter, according to the results previously published by other authors. Two of the synthetic promoters have shown significantly higher activity in endosperm in comparison with GluB4 (+50%). One of the synthetic promoter has determined a two-fold expression level as compared to GluB4 (+113%). Since the reporter protein is phytotoxic, a loss in spatial or temporal expression specificity is supposed to be easily revealed. There has been no evidence of a massive loss of endosperm-specificity, since almost all the rice plants have achieved full maturation and the seed productivity has been enough for analysing a statistically significant number of individuals. Nevertheless, the seed productivity has not been even among the groups. The strongest synthetic promoter has also shown the lowest average seed yield and a high percentage of floral abortion. In order to investigate whether there is a correlation between the incidence of floral abortion and a little loss of spatial or temporal tissue-specificity, the expression level of the reporter gene has been determined in various tissues. A slightly higher content of GCase has been detected in the leaves of the plants transformed with the best synthetic promoter, but the difference with natural promoters has not been statistically significant, also due to the high variability among the individuals of single populations. Among the populations, the average expression level in leaf has been generally proportional to that in endosperm. However, within single populations, there is no correlation either between expression level in leaf and in endosperm or between expression level in leaf and seed yield. No significant difference has been found in the expression of the reporter gene in tissues other than leaf between natural and synthetic promoters. The distribution of the reporter protein between endosperm and aleuronic layer has been almost the same for all natural and synthetic promoters. All these data taken together show that the endosperm specificity had been preserved. The main target of the study has been achieved, since three synthetic promoters have increased the yield of recombinat protein compared to natural RSSP promoters. The iterative method has allowed getting that result without constructing and testing in vivo a large number of synthetic promoters. Furthermore, the methodological approach followed in this work has allowed the identification, within the RSSP promoters, of the regions which are more likely involved in increasing the expression level of the transgene in rice endosperm. For instance: Glb and GluB4 had been found to be good bases for the design of the next synthetic promoters; the 400bp region upstream of the TATA-box of both promoters is particularly rich of cis-elements that drive high transgene expression levels in the endosperm; the Glb portion comprised between the 5’ end and the 400bp downstream region, when fused to the 5' end of the GluB4, has been able to double the yield of the recombinant protein. That portion contains important binding motifs for the RISBZ-2 transcription factor. The importance of those cis-acting elements had been previously underlined by other authors, even though these elements had been reported not to improve the expression level when fused to other RSSP promoters, unless RISBZ-2 has been simultaneously overexpressed. The outcomes of this study have demonstrated, discording with what previously published, that the RISBZ-2 binding region of Glb can be used to successfully improve the expression level of a natural RSSP promoter; also the Prol16 region comprised between the 5’ end and the 500bp downstream region has shown to positively affect the expression level of the transgene, while the results of this PhD advice against the use of the 5' region of GluB1 and GluC in designing synthetic promoters. The outcomes of the sequence analysis have been generally confirmed by the in vivo results. That means the analysis method is consistent and can be exploited to predict the presence of cis-elements. Shuffling these elements in novel, even stronger synthetic endosperm-specific promoters may be the way toward the optimization of an industrial, rice-based, protein-production system.