A dissertation submitted to the Faculty of Science, University of the Witwatersrand, in fulfillment of the requirements for the degree of Master of Science in the School of Molecular and Cell Biology. Johannesburg 2017 / With growing populations and climate change associated drought predicted for the future, cassava can provide one solution for food security and a source of starch for industrial use and biofuels in South Africa, and other countries in the SADC region. One of the severe constraints on cassava production is cassava mosaic disease (CMD) caused by cassava infecting begomoviruse species, including African cassava mosaic virus (ACMV), South African cassava mosaic virus (SACMV) and East African cassava mosaic virus (EACMV). Cassava begomoviruses (CBVs) are responsible for significant yield loss of the starchy tubers. Since no chemical control of virus diseases of plants is possible, one approach to develop virus resistance is via biotechnology, through genetic engineering (GE) of cassava with hairpin RNA (hpRNA) silencing constructs that express small interfering RNAs targeting CBVs and preventing severe disease development. The aim of this project was to subject previously transformed five CMM6 cassava lines (cv. 60444 transformed with a non-mismatched Africa cassava mosaic virus-[Nigeria:Ogorocco;1990] (ACMV-[NG:Ogo:90])-derived hpRNA construct, six AMM2 (cv. 60444 transformed with a mismatched ACMV-[NG:Ogo:90]-derived hpRNA construct), six CMM8 cassava lines (cv.60444 transformed with a non-mismatched SACMV BC1-derived hpRNA construct) and seven AMM4 cassava lines (cv.604444 transformed with a mismatched SACMV BC1-derived hpRNA construct) to reproducible trials, and evaluate for response to virus challenge. The ACMV-[NG:Ogo:90] hpRNAi constructs target 4 overlapping virus open reading frames (ORFs) (AC1 replication associated protein/AC4 and AC2 transcriptional/AC3 replication enhancer), while the SACMV hpRNAi constructs target the cell-to cell movement BC1 ORF. Non mismatched constructs consist of a transformation cassette that has an intron separating the sense and antisense arms of the viral transgene whilst mismatched constructs have the sense arm of the viral transgene treated with bisulfite to induce base mutation. This mutated sense arm is then separated from the non mutated antisense arm by a small spacer. Furthermore, a 229 bp inverted repeat hpRNA construct (DM-AES) was designed to target ACMV-[NG:Ogo:90] 117 nt putative promoter region (2714-49 nt), a 91 nt overlapping sequence (1530-1620 nt) between ACMV-[NG:Ogo:90] AC1 3’ end and AC2 5’ end (AC1 3’/AC2 5’-ter) as well as being efficient against SACMV and EACMV due to the inclusion of a 21 nt conserved sequence (1970-1990) of AC1/Rep shared between ACMV, EACMV and SACMV. Cassava landrace T200 friable embryogenic callus (FEC) were transformed with this construct. The selected transgenic lines were infected with either ACMV-[NG:Ogo:90] (CMM6 and AMM2 transgenic lines) or SACMV (CMM8 and AMM4 transgenic lines) by agro-inoculation and monitored at 14, 36 and 56, 180 and 365 days post infection (dpi) for symptom development, plant growth and viral load. From the ACMV trials 3 lines (CMM6-2, CMM6-6 and line AMM2-52) showed significantly lower symptom scores and lower viral load at 36, 56 and 365 dpi, compared with viral challenged untransgenic cv.60444. This phenotype is described as tolerance, not resistance, as despite ameleriorated symptoms virus replication persists at lower levels. From the SACMV infectivity trials even though all CMM8 and AMM4 transgenic lines had lower symptom severities and viral loads compared with infected untransformed cv.60444, the results were not highly significant (p˃ 0.05). From this study, tolerance or reduction of viral load and symptoms was attributed to the accumulation of transgene-derived siRNAs prior to infection. However there was no observable correlation between levels (semi-qauntitative northern blots) of siRNAs and tolerance or susceptible phenotypes. Tuber yield evaluation of the three tolerant lines (CMM6-2, CMM6-6 and line AMM2-52) showed that the tuber fresh and dry weight at 365 dpi was not affected by the viral presence. These are promising lines for larger greenhouse and field trials. A comparison between the two different constructs showed that the two tolerant CMM6 lines-2 and 6 appeared to perform better (viral load) compared with AMM2 tolerant line-52 with regards to levels of viral amplification. The mismatched construct in AMM4 lines and the nonmismatched construct in CMM8 lines induced the same viral and symptom severity score (sss) reduction. Transformation of T200 FECs with the DM-AES construct was unsuccessful due to the age (more than six months old) of the FECs. FECs are more likely to lose their regeneration and totipotent nature with age. We therefore propose the use of fresh T200 FECs in future transformation studies to test the DM-AES construct. / MT2017