Geminiviruses causes diseases to many staple food and cash crops of great economic importance worldwide. Currently eight species of Begomoviruses belonging to the Geminivirus family exist, of which South African cassava mosaic virus SACMV-[ZA:99] is a member, and is known to cause cassava mosaic disease (CMD). Cassava (Manihot esculenta, Crantz) is considered to be an important food crop consumed in many tropical, sub-tropical and African countries, and is increasingly becoming well-known for its ethanol production on a global a scale. Various strategies to control CMD are currently being implemented, one of which is to elucidate mechanisms involved in host-virus interactions with the aim of identifying defence-related genes involved in the disease process. Many defence genes within the plant kingdom are evolutionary conserved, potentially providing methods of control not only to CMD but to other diseases as well. The research outlined in this thesis aimed to identify networks and pathways involved in disease susceptibility between the model plant host system, Arabidopsis thaliana and cassava T200 upon SACMV-[ZA:99] infection. Conclusions were also drawn from within host comparisons between susceptible cassava T200 and resistant cassava TME3 cultivars in order to explore if similarities, differences or common patterns of expression existed between genes governing resistance and susceptibility. Before transcriptomic profiling studies were carried out, it was important to improve South African cassava mosaic virus (SACMV-[ZA:99]) and African cassava mosaic virus (ACMV-[NG:Ogo:90]) infection efficiencies in recalcitrant crop systems such as cassava. Susceptible cassava cultivars T200, TMS60444, and SM14334 were tested for these purposes following infection with three different Agrobacterium strains (C58C1; AGL1; LBA4404). Results demonstrated that an overall increase in infection efficiency was achieved for each genotype and virus tested, although with varying infectivity levels, suggesting that although an improved method was established, basal levels of susceptibility differed between genotypes and therefore it was not possible to achieve 100% infection efficiencies for agroinfection methods. A 4 x 44k microarray whole genome study was then conducted to identify susceptible host genes involved in the interaction between the model plant system Arabidopsis thaliana and SACMV-[ZA:99]. An infectivity assay was carried out across three time points (14, 24, and 36 dpi), confirming that disease symptoms and virus infectivity levels correlated with an increase in differentially expressed transcripts across time points, with SACMV-[ZA:99] predominantly causing host-gene suppression. Many complex genes and pathways were disrupted and were shown to be involved in categories pertaining to stress and defence responses, phytohormone signalling pathways, cellular transport, metabolism and cell-cycle regulation strongly suggesting an attempt made by SACMV-[ZA:99] to affect homeostasis and antagonize host defence responses. This was the first geminivirus study identifying differentially expressed transcripts across 3 time points. Next generation sequencing (NGS) using the ABI Solid platform was then carried out on SACMV-[ZA:99] – infected susceptible cassava T200 cultivar at 3 time points (12, 32, and 67 dpi), comparing infection responses to mock-inoculated healthy controls. Similarly to the Arabidopsis microarray study, findings from this analysis also revealed a shift from up-regulated to down-regulated genes across time points, once again reflecting virus-specific suppression on host genes suggesting SACMV-[ZA:99] specific alterations were induced in the host, regardless of the host (Arabidopsis and cassava T200) or platform (microarray and NGS) used. Genes identified pertaining particularly to the susceptible cassava T200 - SACMV-[ZA:99] interaction such as the disease resistance protein families (TIR-NBS-LRR), RPP1, RPM1, and NHO1 were showing down-regulation demonstrating that SACMV-[ZA:99] pathogenicity proteins may be causing this suppression leading to inactivation of basal immunity. Comparisons between tolerant cassava TME3 and susceptible T200 showed similarities and differences in responses between the cultivars. Many similarities such as cell wall precursor proteins and glutathione-S-transferases were up-regulated in both cultivars, which may be due to the host attempting to mount appropriate defences. Opposite patterns of expression was observed for genes in categories involved in transcription and phytohormone signalling such as WRKY‘s, NAC, JAZ, and ERF where suppression was evident in susceptible cassava T200, confirming the suppressive nature of SACMV-[ZA:99] to establish a replication-competent environment. Findings in this study contributed to the little that is known about geminivirus disease progression within a previously uncharacterised susceptible host such as cassava.

Cassava (Manihot esculenta Crantz) is affected by two major viral diseases, namely Cassava brown streak disease (CBSD) and Cassava mosaic disease (CMD). Accurate and efficient detection and identification of plant viruses are fundamental aspects of virus diagnosis leading to sustainable disease management. In the present study I describe two techniques, the first based on a single tube duplex and multiplex polymerase chain reaction (m-PCR), developed for simultaneous detection of African cassava mosaic virus (ACMV), East African cassava mosaic Cameroon virus (EACMCV) and East African cassava mosaic Malawi virus (EACMMV), and second, a technique based on Restriction Fragment Length Polymorphism (RFLP) analysis of Reverse Transcribed (RT) -PCR amplified Cassava brown streak viruses species, Cassava brown streak virus (CBSV) and Cassava brown streak Uganda virus (CBSUV). In this work, the single tube duplex and multiplex PCR for simultaneous detection of the four cassava mosaic begomoviruses (CMBs) was developed successfully. Four primer pairs were designed from published DNA-A component sequences targeting specific amplification of the four cassava mosaic begomoviruses (CMBs). Evaluation of the primers sensitivity in serially diluted virus samples revealed that the new primers amplified their target virus to a dilution of 10-4 and 10-3 for uniplex and multiplex PCR respectively. Developed multiplex assay enabled specific amplification of the viruses in producing 950, 503, 435 and 260 base pairs (bp) for ACMV, EACMMV, EACMCV and EACMZV respectively in single and mixed infections of CBSVs. Analysis of 172 field samples from Kenya, Malawi, Mozambique, Rwanda, Tanzania and Zambia detected both single and mixed infections, results which were proved by analysis of the sequenced amplicons. Second, a technique based on 2 Restriction Fragment Length Polymorphism (RFLP) analysis of RT-PCR amplified cassava brown streak viruses, Cassava brown streak virus (CBSV) and cassava brown streak Uganda virus (CBSUV), was performed. A degenerate primer amplifying 785 bp of the coat protein gene (CP) of CBSV and CBSUV was also designed. Two restriction endonucleases, HindIII and EcoR1 (identified by a software package, Vector NTI® Express v1.0 from Life Technologies/Invitrogen), which produce different fragments upon digestion of RT-PCR amplicons from CBSV and CBSUV, were used to distinguish the two viruses RFLP analysis using EcoRI has no site in CBSV producing one fragment (785 bp), two fragments (525 bp and 224 bp) for CBSUV and three fragments (785, 525 and 224 bp) for the mixed infections. On the other hand, HindIII has no site in CBSUV producing one fragment (785 bp), three fragments (437 bp, 267 bp and 81 bp) were produced for CBSV, and four fragments (785, 437, 267 and 81 bp) for CBSV and CBSUV mixed infections. In both multiplex and RFLP analyses, results from the sequenced PCR/RT-PCR amplicons agreed with sequence identities of the respective published virus species. Experience from using developed multiplex and RFLP techniques show that time was saved and amount of reagents used were reduced. RFLPs confirmed the presence of CBSV and CBSUV in RT-PCR amplicons without requirement for sequencing. Additionally, modified protocols from Dellaporta et al. (1983) and Chang et al. (1993), were used to extract DNA and RNA respectively from dry and fresh cassava leaves with comparable results. I also demonstrated a method of collecting and preserving cassava leaf samples to retain their integrity during storage for a period of over one month. The two diagnostic tools can be used routinely in germplasm indexing, disease surveillance, and disease monitoring programs 3 Problem Statement and Rationale In east and southern Africa, cassava (Manihot esculenta Crantz) is one of the leading crops in terms of production and has become an important source of income to households and small-scale farmers. However, the production across the region is greatly affected by Cassava mosaic disease (CMD) and Cassava brown streak disease (CBSD). Reports from different authors (Gibson. 1996; Ogbe et al., 1996; Legg et al., 1999; Fondong et al., 2000; Bisimwa et al., 2012) have reported the occurrence of CMD in different countries in the SSA. In Tanzania, CMD has been reported from many locations. Comprehensive characterization by Ndunguru et al. (2005) showed seven cassava mosaic geminiviruses species occur in Tanzania. Mbanzibwa et al. (2009a) reported prevalence of two potyvirus species causing CBSD in the Lake Victoria basin and along the coastal belt of Indian Ocean. A countrywide survey of all major cassava-growing areas in Kenya by Bull et al. (2006) reported presence of six CMG species with novel begomoviruses and a new recombinant strain of EACMV, demonstrating increasing diversity and geographical distribution of CMGs. Similarly, recent reemergence of CBSD has been reported in many districts in Uganda (Alicai et al., 2007) as well as from Malawi (Winter et al., 2010), Kenya (Mware et al., 2009) and Rwanda (Shirima et al., 2012). No reports of occurrence of CBSD have been reported from Zambia. With the current development of more robust diagnostic tools such as RT-PCR and real-time PCR, the diagnosis of CMD and CBSD has also improved in many cassava- producing countries. Similarly, the challenges to obtain more sensitive broad-spectrum cost-effective diagnostic tools also increase. This is evident following discovery of more 4 virus species causing CMD and CBSD (Mbanzibwa et al., 2009a and Winter et al., 2010) which can easily be overlooked. In the field the co-infections of many CMBs and CBSVs is common. Therefore, it will require several tools to detect the multiple infections using the diagnostic tools currently available. Thus, development of efficient and affordable diagnostic tools for simultaneous detection and identification of CMBs and CBSVs is vital and will have a significant impact on development and implementation of cassava virus disease management. Diagnostics will be used for disease monitoring in cassava multiplication plots production and distribution of disease- free cassava planting materials. Therefore, this research make use of the available sequence information in the database for both CMBs and CBSVs to develop sensitive tools for the simultaneous detection of four species of cassava begomoviruses namely: African cassava mosaic virus (ACMV), East African cassava mosaic Cameroon virus (EACMCV), East African cassava mosaic Malawi virus (EACMMV) and East African cassava Mosaic Zanzibar Virus (EACMZV) using multiplex PCR. Also identification and differentiation of two species of Cassava brown streak viruses namely Cassava brown streak virus (CBSV) and Cassava brown streak Uganda virus (CBSUV) by RT-PCR/RFLP approach. This study generated knowledge and new tools that will enhance the diagnosis of both CMD and CBSD. The tools will facilitate deployment of virus-indexed cassava planting materials within the region.

MSc (Plant Production) / Department of Plant Production / See the attached abstract below